diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.6068801160501502,
-  "best_model_checkpoint": "/m/triton/scratch/elec/puhe/p/palp3/MUCS/mucs_language_segregated_data/trainwithtags_warmup1500_s300_shuff100/output/checkpoint-1000",
-  "epoch": 5.6,
-  "eval_steps": 100,
-  "global_step": 3500,
+  "best_metric": 0.6194798466480157,
+  "best_model_checkpoint": "/m/triton/scratch/elec/puhe/p/palp3/MUCS/mucs_language_segregated_data/trainwithtags_warmup2000_s300_shuff100/output/checkpoint-1000",
+  "epoch": 9.6,
+  "eval_steps": 1000,
+  "global_step": 6000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -17,24862 +17,42072 @@
     },
     {
       "epoch": 0.0032,
-      "grad_norm": 19.480770111083984,
-      "learning_rate": 3.9999999999999993e-07,
+      "grad_norm": 19.478151321411133,
+      "learning_rate": 3e-07,
       "loss": 28.9967,
       "step": 2
     },
     {
       "epoch": 0.0048,
-      "grad_norm": 21.03929901123047,
-      "learning_rate": 7.999999999999999e-07,
+      "grad_norm": 21.032878875732422,
+      "learning_rate": 6e-07,
       "loss": 25.1809,
       "step": 3
     },
     {
       "epoch": 0.0064,
-      "grad_norm": 13.241178512573242,
-      "learning_rate": 1.2e-06,
+      "grad_norm": 13.343757629394531,
+      "learning_rate": 9e-07,
       "loss": 20.6383,
       "step": 4
     },
     {
       "epoch": 0.008,
       "grad_norm": NaN,
-      "learning_rate": 1.2e-06,
-      "loss": 19.1797,
+      "learning_rate": 9e-07,
+      "loss": 19.1799,
       "step": 5
     },
     {
       "epoch": 0.0096,
-      "grad_norm": 13.427491188049316,
-      "learning_rate": 1.5999999999999997e-06,
+      "grad_norm": 12.849465370178223,
+      "learning_rate": 1.2e-06,
       "loss": 20.0215,
       "step": 6
     },
     {
       "epoch": 0.0112,
-      "grad_norm": 13.29790210723877,
-      "learning_rate": 2e-06,
-      "loss": 20.0473,
+      "grad_norm": 13.198436737060547,
+      "learning_rate": 1.4999999999999998e-06,
+      "loss": 20.0485,
       "step": 7
     },
     {
       "epoch": 0.0128,
-      "grad_norm": 11.687248229980469,
-      "learning_rate": 2.4e-06,
-      "loss": 18.6201,
+      "grad_norm": 11.687944412231445,
+      "learning_rate": 1.8e-06,
+      "loss": 18.6217,
       "step": 8
     },
     {
       "epoch": 0.0144,
-      "grad_norm": 10.248677253723145,
-      "learning_rate": 2.8e-06,
-      "loss": 16.4545,
+      "grad_norm": 10.567115783691406,
+      "learning_rate": 2.1e-06,
+      "loss": 16.4561,
       "step": 9
     },
     {
       "epoch": 0.016,
-      "grad_norm": 10.466910362243652,
-      "learning_rate": 3.1999999999999994e-06,
-      "loss": 16.9539,
+      "grad_norm": 10.45132064819336,
+      "learning_rate": 2.4e-06,
+      "loss": 16.9583,
       "step": 10
     },
     {
       "epoch": 0.0176,
-      "grad_norm": 13.025243759155273,
-      "learning_rate": 3.6e-06,
-      "loss": 17.2907,
+      "grad_norm": 12.070805549621582,
+      "learning_rate": 2.6999999999999996e-06,
+      "loss": 17.2946,
       "step": 11
     },
     {
       "epoch": 0.0192,
-      "grad_norm": 10.613122940063477,
-      "learning_rate": 4e-06,
-      "loss": 17.1058,
+      "grad_norm": 10.647071838378906,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 17.1101,
       "step": 12
     },
     {
       "epoch": 0.0208,
-      "grad_norm": 9.954842567443848,
-      "learning_rate": 4.399999999999999e-06,
-      "loss": 16.4296,
+      "grad_norm": 9.95898723602295,
+      "learning_rate": 3.2999999999999993e-06,
+      "loss": 16.4421,
       "step": 13
     },
     {
       "epoch": 0.0224,
-      "grad_norm": 11.088004112243652,
-      "learning_rate": 4.8e-06,
-      "loss": 17.5867,
+      "grad_norm": 11.099569320678711,
+      "learning_rate": 3.6e-06,
+      "loss": 17.5943,
       "step": 14
     },
     {
       "epoch": 0.024,
-      "grad_norm": 10.448060989379883,
-      "learning_rate": 5.199999999999999e-06,
-      "loss": 16.871,
+      "grad_norm": 10.42167854309082,
+      "learning_rate": 3.899999999999999e-06,
+      "loss": 16.8807,
       "step": 15
     },
     {
       "epoch": 0.0256,
-      "grad_norm": 8.993600845336914,
-      "learning_rate": 5.6e-06,
-      "loss": 14.601,
+      "grad_norm": 9.058366775512695,
+      "learning_rate": 4.2e-06,
+      "loss": 14.6257,
       "step": 16
     },
     {
       "epoch": 0.0272,
-      "grad_norm": 8.905482292175293,
-      "learning_rate": 5.999999999999999e-06,
-      "loss": 14.0951,
+      "grad_norm": 9.066213607788086,
+      "learning_rate": 4.499999999999999e-06,
+      "loss": 14.1237,
       "step": 17
     },
     {
       "epoch": 0.0288,
-      "grad_norm": 9.374123573303223,
-      "learning_rate": 6.399999999999999e-06,
-      "loss": 15.2176,
+      "grad_norm": 9.31420612335205,
+      "learning_rate": 4.8e-06,
+      "loss": 15.2356,
       "step": 18
     },
     {
       "epoch": 0.0304,
-      "grad_norm": 19.181642532348633,
-      "learning_rate": 6.8e-06,
-      "loss": 15.4023,
+      "grad_norm": 10.176087379455566,
+      "learning_rate": 5.1e-06,
+      "loss": 15.4251,
       "step": 19
     },
     {
       "epoch": 0.032,
-      "grad_norm": 8.2308988571167,
-      "learning_rate": 7.2e-06,
-      "loss": 13.7388,
+      "grad_norm": 8.170927047729492,
+      "learning_rate": 5.399999999999999e-06,
+      "loss": 13.76,
       "step": 20
     },
     {
       "epoch": 0.0336,
-      "grad_norm": 8.79343032836914,
-      "learning_rate": 7.599999999999999e-06,
-      "loss": 14.4443,
+      "grad_norm": 8.670867919921875,
+      "learning_rate": 5.7e-06,
+      "loss": 14.4806,
       "step": 21
     },
     {
       "epoch": 0.0352,
-      "grad_norm": 11.329385757446289,
-      "learning_rate": 8e-06,
-      "loss": 16.5094,
+      "grad_norm": 11.31945514678955,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 16.5971,
       "step": 22
     },
     {
       "epoch": 0.0368,
-      "grad_norm": 9.5508451461792,
-      "learning_rate": 8.4e-06,
-      "loss": 14.956,
+      "grad_norm": 9.18188762664795,
+      "learning_rate": 6.3e-06,
+      "loss": 15.0373,
       "step": 23
     },
     {
       "epoch": 0.0384,
-      "grad_norm": 9.614250183105469,
-      "learning_rate": 8.799999999999999e-06,
-      "loss": 14.7454,
+      "grad_norm": 9.432085990905762,
+      "learning_rate": 6.599999999999999e-06,
+      "loss": 14.792,
       "step": 24
     },
     {
       "epoch": 0.04,
-      "grad_norm": 9.611518859863281,
-      "learning_rate": 9.199999999999998e-06,
-      "loss": 14.8943,
+      "grad_norm": 9.444454193115234,
+      "learning_rate": 6.899999999999999e-06,
+      "loss": 14.9425,
       "step": 25
     },
     {
       "epoch": 0.0416,
-      "grad_norm": 10.2857084274292,
-      "learning_rate": 9.6e-06,
-      "loss": 15.5331,
+      "grad_norm": 10.104069709777832,
+      "learning_rate": 7.2e-06,
+      "loss": 15.5962,
       "step": 26
     },
     {
       "epoch": 0.0432,
-      "grad_norm": 8.22513198852539,
-      "learning_rate": 9.999999999999999e-06,
-      "loss": 13.0081,
+      "grad_norm": 8.144676208496094,
+      "learning_rate": 7.499999999999999e-06,
+      "loss": 13.0629,
       "step": 27
     },
     {
       "epoch": 0.0448,
-      "grad_norm": 11.55008602142334,
-      "learning_rate": 1.0399999999999999e-05,
-      "loss": 16.8155,
+      "grad_norm": 11.250031471252441,
+      "learning_rate": 7.799999999999998e-06,
+      "loss": 16.8965,
       "step": 28
     },
     {
       "epoch": 0.0464,
-      "grad_norm": 9.69379711151123,
-      "learning_rate": 1.0799999999999998e-05,
-      "loss": 14.1567,
+      "grad_norm": 9.27499008178711,
+      "learning_rate": 8.099999999999999e-06,
+      "loss": 14.2719,
       "step": 29
     },
     {
       "epoch": 0.048,
-      "grad_norm": 9.33486557006836,
-      "learning_rate": 1.12e-05,
-      "loss": 13.6537,
+      "grad_norm": 8.996031761169434,
+      "learning_rate": 8.4e-06,
+      "loss": 13.7765,
       "step": 30
     },
     {
       "epoch": 0.0496,
-      "grad_norm": 10.678018569946289,
-      "learning_rate": 1.1599999999999999e-05,
-      "loss": 15.083,
+      "grad_norm": 10.058349609375,
+      "learning_rate": 8.7e-06,
+      "loss": 15.2312,
       "step": 31
     },
     {
       "epoch": 0.0512,
-      "grad_norm": 10.413251876831055,
-      "learning_rate": 1.1999999999999999e-05,
-      "loss": 14.4103,
+      "grad_norm": 9.924124717712402,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 14.5462,
       "step": 32
     },
     {
       "epoch": 0.0528,
-      "grad_norm": 11.2171630859375,
-      "learning_rate": 1.2399999999999998e-05,
-      "loss": 15.0434,
+      "grad_norm": 10.69217586517334,
+      "learning_rate": 9.299999999999999e-06,
+      "loss": 15.1555,
       "step": 33
     },
     {
       "epoch": 0.0544,
-      "grad_norm": 8.697986602783203,
-      "learning_rate": 1.2799999999999998e-05,
-      "loss": 12.6689,
+      "grad_norm": 8.319705963134766,
+      "learning_rate": 9.6e-06,
+      "loss": 12.7821,
       "step": 34
     },
     {
       "epoch": 0.056,
-      "grad_norm": 10.12214183807373,
-      "learning_rate": 1.3199999999999997e-05,
-      "loss": 13.8483,
+      "grad_norm": 9.608613967895508,
+      "learning_rate": 9.9e-06,
+      "loss": 13.9979,
       "step": 35
     },
     {
       "epoch": 0.0576,
-      "grad_norm": 10.093426704406738,
-      "learning_rate": 1.36e-05,
-      "loss": 13.6302,
+      "grad_norm": 9.609368324279785,
+      "learning_rate": 1.02e-05,
+      "loss": 13.7818,
       "step": 36
     },
     {
       "epoch": 0.0592,
-      "grad_norm": 10.467079162597656,
-      "learning_rate": 1.4e-05,
-      "loss": 13.8487,
+      "grad_norm": 9.907469749450684,
+      "learning_rate": 1.05e-05,
+      "loss": 14.0173,
       "step": 37
     },
     {
       "epoch": 0.0608,
-      "grad_norm": 11.031571388244629,
-      "learning_rate": 1.44e-05,
-      "loss": 14.0714,
+      "grad_norm": 10.462899208068848,
+      "learning_rate": 1.0799999999999998e-05,
+      "loss": 14.2503,
       "step": 38
     },
     {
       "epoch": 0.0624,
-      "grad_norm": 11.238005638122559,
-      "learning_rate": 1.4799999999999999e-05,
-      "loss": 14.0722,
+      "grad_norm": 10.519027709960938,
+      "learning_rate": 1.1099999999999999e-05,
+      "loss": 14.2442,
       "step": 39
     },
     {
       "epoch": 0.064,
-      "grad_norm": 11.471139907836914,
-      "learning_rate": 1.5199999999999998e-05,
-      "loss": 13.9102,
+      "grad_norm": 10.643707275390625,
+      "learning_rate": 1.14e-05,
+      "loss": 14.0883,
       "step": 40
     },
     {
       "epoch": 0.0656,
-      "grad_norm": 13.19694995880127,
-      "learning_rate": 1.5599999999999996e-05,
-      "loss": 15.5816,
+      "grad_norm": 12.463756561279297,
+      "learning_rate": 1.17e-05,
+      "loss": 15.8071,
       "step": 41
     },
     {
       "epoch": 0.0672,
-      "grad_norm": 15.062193870544434,
-      "learning_rate": 1.6e-05,
-      "loss": 16.0428,
+      "grad_norm": 13.493027687072754,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 16.2933,
       "step": 42
     },
     {
       "epoch": 0.0688,
-      "grad_norm": 14.18619155883789,
-      "learning_rate": 1.64e-05,
-      "loss": 15.99,
+      "grad_norm": 13.236719131469727,
+      "learning_rate": 1.2299999999999999e-05,
+      "loss": 16.2827,
       "step": 43
     },
     {
       "epoch": 0.0704,
-      "grad_norm": 11.722272872924805,
-      "learning_rate": 1.68e-05,
-      "loss": 13.8123,
+      "grad_norm": 10.915094375610352,
+      "learning_rate": 1.26e-05,
+      "loss": 14.0712,
       "step": 44
     },
     {
       "epoch": 0.072,
-      "grad_norm": 12.653892517089844,
-      "learning_rate": 1.7199999999999998e-05,
-      "loss": 14.2928,
+      "grad_norm": 11.671866416931152,
+      "learning_rate": 1.2899999999999998e-05,
+      "loss": 14.5855,
       "step": 45
     },
     {
       "epoch": 0.0736,
-      "grad_norm": 12.380364418029785,
-      "learning_rate": 1.7599999999999998e-05,
-      "loss": 13.9655,
+      "grad_norm": 11.406953811645508,
+      "learning_rate": 1.3199999999999997e-05,
+      "loss": 14.2818,
       "step": 46
     },
     {
       "epoch": 0.0752,
-      "grad_norm": 10.684913635253906,
-      "learning_rate": 1.7999999999999997e-05,
-      "loss": 12.4269,
+      "grad_norm": 9.801484107971191,
+      "learning_rate": 1.3499999999999998e-05,
+      "loss": 12.7244,
       "step": 47
     },
     {
       "epoch": 0.0768,
-      "grad_norm": 13.753266334533691,
-      "learning_rate": 1.8399999999999997e-05,
-      "loss": 14.3082,
+      "grad_norm": 12.542718887329102,
+      "learning_rate": 1.3799999999999998e-05,
+      "loss": 14.6995,
       "step": 48
     },
     {
       "epoch": 0.0784,
-      "grad_norm": 14.648118019104004,
-      "learning_rate": 1.8799999999999996e-05,
-      "loss": 14.7347,
+      "grad_norm": 13.37395191192627,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 15.1658,
       "step": 49
     },
     {
       "epoch": 0.08,
-      "grad_norm": 12.614858627319336,
-      "learning_rate": 1.92e-05,
-      "loss": 13.0945,
+      "grad_norm": 11.486909866333008,
+      "learning_rate": 1.44e-05,
+      "loss": 13.4777,
       "step": 50
     },
     {
       "epoch": 0.0816,
-      "grad_norm": 46.947227478027344,
-      "learning_rate": 1.9599999999999995e-05,
-      "loss": 30.2031,
+      "grad_norm": 41.783447265625,
+      "learning_rate": 1.47e-05,
+      "loss": 31.3796,
       "step": 51
     },
     {
       "epoch": 0.0832,
-      "grad_norm": NaN,
-      "learning_rate": 1.9599999999999995e-05,
-      "loss": 24.7875,
+      "grad_norm": 37.17393112182617,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 25.7284,
       "step": 52
     },
     {
       "epoch": 0.0848,
-      "grad_norm": 32.876708984375,
-      "learning_rate": 1.9999999999999998e-05,
-      "loss": 25.5323,
+      "grad_norm": 30.743728637695312,
+      "learning_rate": 1.53e-05,
+      "loss": 26.3766,
       "step": 53
     },
     {
       "epoch": 0.0864,
-      "grad_norm": 22.72123146057129,
-      "learning_rate": 2.04e-05,
-      "loss": 18.7279,
+      "grad_norm": 21.993959426879883,
+      "learning_rate": 1.5599999999999996e-05,
+      "loss": 19.3039,
       "step": 54
     },
     {
       "epoch": 0.088,
-      "grad_norm": 60.50054168701172,
-      "learning_rate": 2.0799999999999997e-05,
-      "loss": 25.0974,
+      "grad_norm": NaN,
+      "learning_rate": 1.5599999999999996e-05,
+      "loss": 25.9385,
       "step": 55
     },
     {
       "epoch": 0.0896,
-      "grad_norm": 19.52132225036621,
-      "learning_rate": 2.12e-05,
-      "loss": 16.16,
+      "grad_norm": 17.90077781677246,
+      "learning_rate": 1.5899999999999997e-05,
+      "loss": 16.7533,
       "step": 56
     },
     {
       "epoch": 0.0912,
-      "grad_norm": 33.881954193115234,
-      "learning_rate": 2.1599999999999996e-05,
-      "loss": 23.8675,
+      "grad_norm": 30.298683166503906,
+      "learning_rate": 1.6199999999999997e-05,
+      "loss": 24.8963,
       "step": 57
     },
     {
       "epoch": 0.0928,
-      "grad_norm": 18.729801177978516,
-      "learning_rate": 2.2e-05,
-      "loss": 15.2585,
+      "grad_norm": 16.790489196777344,
+      "learning_rate": 1.6499999999999998e-05,
+      "loss": 15.8818,
       "step": 58
     },
     {
       "epoch": 0.0944,
-      "grad_norm": 18.455617904663086,
-      "learning_rate": 2.24e-05,
-      "loss": 14.5196,
+      "grad_norm": 16.33099937438965,
+      "learning_rate": 1.68e-05,
+      "loss": 15.1584,
       "step": 59
     },
     {
       "epoch": 0.096,
-      "grad_norm": 19.266185760498047,
-      "learning_rate": 2.28e-05,
-      "loss": 15.0203,
+      "grad_norm": 17.538009643554688,
+      "learning_rate": 1.71e-05,
+      "loss": 15.7514,
       "step": 60
     },
     {
       "epoch": 0.0976,
-      "grad_norm": 21.54351234436035,
-      "learning_rate": 2.3199999999999998e-05,
-      "loss": 15.7071,
+      "grad_norm": 19.184165954589844,
+      "learning_rate": 1.74e-05,
+      "loss": 16.5568,
       "step": 61
     },
     {
       "epoch": 0.0992,
-      "grad_norm": 18.729368209838867,
-      "learning_rate": 2.3599999999999998e-05,
-      "loss": 13.9821,
+      "grad_norm": 16.525760650634766,
+      "learning_rate": 1.7699999999999997e-05,
+      "loss": 14.7701,
       "step": 62
     },
     {
       "epoch": 0.1008,
-      "grad_norm": 19.222980499267578,
-      "learning_rate": 2.3999999999999997e-05,
-      "loss": 13.5072,
+      "grad_norm": 17.058059692382812,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 14.3009,
       "step": 63
     },
     {
       "epoch": 0.1024,
-      "grad_norm": 18.127918243408203,
-      "learning_rate": 2.4399999999999997e-05,
-      "loss": 13.0847,
+      "grad_norm": 15.610758781433105,
+      "learning_rate": 1.8299999999999998e-05,
+      "loss": 13.9059,
       "step": 64
     },
     {
       "epoch": 0.104,
-      "grad_norm": 17.083431243896484,
-      "learning_rate": 2.4799999999999996e-05,
-      "loss": 12.3284,
+      "grad_norm": 14.720325469970703,
+      "learning_rate": 1.8599999999999998e-05,
+      "loss": 13.1319,
       "step": 65
     },
     {
       "epoch": 0.1056,
-      "grad_norm": 20.064725875854492,
-      "learning_rate": 2.52e-05,
-      "loss": 13.3525,
+      "grad_norm": 17.151647567749023,
+      "learning_rate": 1.89e-05,
+      "loss": 14.3394,
       "step": 66
     },
     {
       "epoch": 0.1072,
-      "grad_norm": 19.82396697998047,
-      "learning_rate": 2.5599999999999995e-05,
-      "loss": 12.8442,
+      "grad_norm": 16.752134323120117,
+      "learning_rate": 1.92e-05,
+      "loss": 13.8388,
       "step": 67
     },
     {
       "epoch": 0.1088,
-      "grad_norm": 19.43636131286621,
-      "learning_rate": 2.6e-05,
-      "loss": 12.5456,
+      "grad_norm": 16.487157821655273,
+      "learning_rate": 1.95e-05,
+      "loss": 13.5623,
       "step": 68
     },
     {
       "epoch": 0.1104,
-      "grad_norm": 19.312517166137695,
-      "learning_rate": 2.6399999999999995e-05,
-      "loss": 12.0877,
+      "grad_norm": 16.15814781188965,
+      "learning_rate": 1.98e-05,
+      "loss": 13.1342,
       "step": 69
     },
     {
       "epoch": 0.112,
-      "grad_norm": 19.31719970703125,
-      "learning_rate": 2.6799999999999998e-05,
-      "loss": 11.4867,
+      "grad_norm": 15.702254295349121,
+      "learning_rate": 2.01e-05,
+      "loss": 12.5287,
       "step": 70
     },
     {
       "epoch": 0.1136,
-      "grad_norm": 20.121965408325195,
-      "learning_rate": 2.72e-05,
-      "loss": 11.5062,
+      "grad_norm": 16.827510833740234,
+      "learning_rate": 2.04e-05,
+      "loss": 12.6514,
       "step": 71
     },
     {
       "epoch": 0.1152,
-      "grad_norm": 19.37114715576172,
-      "learning_rate": 2.7599999999999997e-05,
-      "loss": 11.0737,
+      "grad_norm": 16.10146713256836,
+      "learning_rate": 2.07e-05,
+      "loss": 12.2281,
       "step": 72
     },
     {
       "epoch": 0.1168,
-      "grad_norm": 21.755847930908203,
-      "learning_rate": 2.8e-05,
-      "loss": 11.4482,
+      "grad_norm": 17.98247528076172,
+      "learning_rate": 2.1e-05,
+      "loss": 12.7926,
       "step": 73
     },
     {
       "epoch": 0.1184,
-      "grad_norm": 22.922266006469727,
-      "learning_rate": 2.8399999999999996e-05,
-      "loss": 11.6379,
+      "grad_norm": 18.706369400024414,
+      "learning_rate": 2.1299999999999996e-05,
+      "loss": 13.0467,
       "step": 74
     },
     {
       "epoch": 0.12,
-      "grad_norm": 25.408145904541016,
-      "learning_rate": 2.88e-05,
-      "loss": 11.4125,
+      "grad_norm": 19.2159423828125,
+      "learning_rate": 2.1599999999999996e-05,
+      "loss": 12.9773,
       "step": 75
     },
     {
       "epoch": 0.1216,
-      "grad_norm": 23.630056381225586,
-      "learning_rate": 2.9199999999999995e-05,
-      "loss": 11.1628,
+      "grad_norm": 18.725379943847656,
+      "learning_rate": 2.1899999999999997e-05,
+      "loss": 12.7265,
       "step": 76
     },
     {
       "epoch": 0.1232,
-      "grad_norm": 24.943918228149414,
-      "learning_rate": 2.9599999999999998e-05,
-      "loss": 10.9194,
+      "grad_norm": 19.681507110595703,
+      "learning_rate": 2.2199999999999998e-05,
+      "loss": 12.6436,
       "step": 77
     },
     {
       "epoch": 0.1248,
-      "grad_norm": 26.104032516479492,
-      "learning_rate": 2.9999999999999997e-05,
-      "loss": 10.7166,
+      "grad_norm": 20.383150100708008,
+      "learning_rate": 2.2499999999999998e-05,
+      "loss": 12.5706,
       "step": 78
     },
     {
       "epoch": 0.1264,
-      "grad_norm": 23.736133575439453,
-      "learning_rate": 3.0399999999999997e-05,
-      "loss": 9.6632,
+      "grad_norm": 18.141130447387695,
+      "learning_rate": 2.28e-05,
+      "loss": 11.3421,
       "step": 79
     },
     {
       "epoch": 0.128,
-      "grad_norm": 29.95717430114746,
-      "learning_rate": 3.0799999999999996e-05,
-      "loss": 10.7768,
+      "grad_norm": 23.76494789123535,
+      "learning_rate": 2.31e-05,
+      "loss": 12.8837,
       "step": 80
     },
     {
       "epoch": 0.1296,
-      "grad_norm": 27.857250213623047,
-      "learning_rate": 3.119999999999999e-05,
-      "loss": 9.604,
+      "grad_norm": 21.412139892578125,
+      "learning_rate": 2.34e-05,
+      "loss": 11.5721,
       "step": 81
     },
     {
       "epoch": 0.1312,
-      "grad_norm": 31.466289520263672,
-      "learning_rate": 3.1599999999999996e-05,
-      "loss": 10.2822,
+      "grad_norm": 24.82501792907715,
+      "learning_rate": 2.3699999999999997e-05,
+      "loss": 12.6466,
       "step": 82
     },
     {
       "epoch": 0.1328,
-      "grad_norm": 25.81043243408203,
-      "learning_rate": 3.2e-05,
-      "loss": 8.937,
+      "grad_norm": 20.29831886291504,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 10.9449,
       "step": 83
     },
     {
       "epoch": 0.1344,
-      "grad_norm": 32.71095275878906,
-      "learning_rate": 3.2399999999999995e-05,
-      "loss": 9.828,
+      "grad_norm": 26.979677200317383,
+      "learning_rate": 2.4299999999999998e-05,
+      "loss": 12.3749,
       "step": 84
     },
     {
       "epoch": 0.136,
-      "grad_norm": 29.731168746948242,
-      "learning_rate": 3.28e-05,
-      "loss": 8.8936,
+      "grad_norm": 23.62879180908203,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 11.1793,
       "step": 85
     },
     {
       "epoch": 0.1376,
-      "grad_norm": 30.395309448242188,
-      "learning_rate": 3.3199999999999994e-05,
-      "loss": 8.8308,
+      "grad_norm": 27.601375579833984,
+      "learning_rate": 2.49e-05,
+      "loss": 11.2504,
       "step": 86
     },
     {
       "epoch": 0.1392,
-      "grad_norm": 31.029926300048828,
-      "learning_rate": 3.36e-05,
-      "loss": 8.7211,
+      "grad_norm": 26.318252563476562,
+      "learning_rate": 2.52e-05,
+      "loss": 11.2328,
       "step": 87
     },
     {
       "epoch": 0.1408,
-      "grad_norm": 28.081674575805664,
-      "learning_rate": 3.399999999999999e-05,
-      "loss": 7.9264,
+      "grad_norm": 24.130613327026367,
+      "learning_rate": 2.55e-05,
+      "loss": 10.2347,
       "step": 88
     },
     {
       "epoch": 0.1424,
-      "grad_norm": 25.76122283935547,
-      "learning_rate": 3.4399999999999996e-05,
-      "loss": 7.4507,
+      "grad_norm": 22.916906356811523,
+      "learning_rate": 2.5799999999999997e-05,
+      "loss": 9.6041,
       "step": 89
     },
     {
       "epoch": 0.144,
-      "grad_norm": 29.275390625,
-      "learning_rate": 3.48e-05,
-      "loss": 7.8001,
+      "grad_norm": 26.457481384277344,
+      "learning_rate": 2.6099999999999997e-05,
+      "loss": 10.3063,
       "step": 90
     },
     {
       "epoch": 0.1456,
-      "grad_norm": 26.79217529296875,
-      "learning_rate": 3.5199999999999995e-05,
-      "loss": 7.3485,
+      "grad_norm": 24.79904556274414,
+      "learning_rate": 2.6399999999999995e-05,
+      "loss": 9.7561,
       "step": 91
     },
     {
       "epoch": 0.1472,
-      "grad_norm": 26.25724220275879,
-      "learning_rate": 3.56e-05,
-      "loss": 7.0524,
+      "grad_norm": 25.292531967163086,
+      "learning_rate": 2.6699999999999995e-05,
+      "loss": 9.4004,
       "step": 92
     },
     {
       "epoch": 0.1488,
-      "grad_norm": 23.24335289001465,
-      "learning_rate": 3.5999999999999994e-05,
-      "loss": 6.5839,
+      "grad_norm": 23.30132484436035,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 8.7521,
       "step": 93
     },
     {
       "epoch": 0.1504,
-      "grad_norm": 22.426977157592773,
-      "learning_rate": 3.64e-05,
-      "loss": 6.4155,
+      "grad_norm": 23.52458381652832,
+      "learning_rate": 2.7299999999999996e-05,
+      "loss": 8.5213,
       "step": 94
     },
     {
       "epoch": 0.152,
-      "grad_norm": 30.2196102142334,
-      "learning_rate": 3.679999999999999e-05,
-      "loss": 6.9315,
+      "grad_norm": 31.594478607177734,
+      "learning_rate": 2.7599999999999997e-05,
+      "loss": 9.6137,
       "step": 95
     },
     {
       "epoch": 0.1536,
-      "grad_norm": 24.466903686523438,
-      "learning_rate": 3.7199999999999996e-05,
-      "loss": 6.1573,
+      "grad_norm": 25.19120979309082,
+      "learning_rate": 2.7899999999999997e-05,
+      "loss": 8.2984,
       "step": 96
     },
     {
       "epoch": 0.1552,
-      "grad_norm": 29.544233322143555,
-      "learning_rate": 3.759999999999999e-05,
-      "loss": 6.8333,
+      "grad_norm": 34.27452850341797,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 9.7313,
       "step": 97
     },
     {
       "epoch": 0.1568,
-      "grad_norm": 15.694954872131348,
-      "learning_rate": 3.8e-05,
-      "loss": 5.5889,
+      "grad_norm": 20.038379669189453,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 7.3048,
       "step": 98
     },
     {
       "epoch": 0.1584,
-      "grad_norm": 18.473628997802734,
-      "learning_rate": 3.84e-05,
-      "loss": 5.7588,
+      "grad_norm": 24.651748657226562,
+      "learning_rate": 2.88e-05,
+      "loss": 7.757,
       "step": 99
     },
     {
       "epoch": 0.16,
-      "grad_norm": 12.698282241821289,
-      "learning_rate": 3.8799999999999994e-05,
-      "loss": 5.2909,
-      "step": 100
-    },
-    {
-      "epoch": 0.16,
-      "eval_cer": 0.9933272212699805,
-      "eval_loss": 5.788337707519531,
-      "eval_runtime": 158.8341,
-      "eval_samples_per_second": 19.744,
-      "eval_steps_per_second": 1.234,
-      "eval_wer": 1.0,
+      "grad_norm": 19.042896270751953,
+      "learning_rate": 2.91e-05,
+      "loss": 6.7935,
       "step": 100
     },
     {
       "epoch": 0.1616,
       "grad_norm": NaN,
-      "learning_rate": 3.8799999999999994e-05,
-      "loss": 9.8624,
+      "learning_rate": 2.91e-05,
+      "loss": 16.7497,
       "step": 101
     },
     {
       "epoch": 0.1632,
-      "grad_norm": 55.51943588256836,
-      "learning_rate": 3.919999999999999e-05,
-      "loss": 6.8633,
+      "grad_norm": 59.27364730834961,
+      "learning_rate": 2.94e-05,
+      "loss": 10.7056,
       "step": 102
     },
     {
       "epoch": 0.1648,
-      "grad_norm": 31.776363372802734,
-      "learning_rate": 3.96e-05,
-      "loss": 5.9656,
+      "grad_norm": 40.90422821044922,
+      "learning_rate": 2.97e-05,
+      "loss": 8.9462,
       "step": 103
     },
     {
       "epoch": 0.1664,
-      "grad_norm": 41.10131072998047,
-      "learning_rate": 3.9999999999999996e-05,
-      "loss": 6.4203,
+      "grad_norm": 95.33076477050781,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 9.9958,
       "step": 104
     },
     {
       "epoch": 0.168,
-      "grad_norm": 22.717975616455078,
-      "learning_rate": 4.039999999999999e-05,
-      "loss": 5.4663,
+      "grad_norm": 34.60516357421875,
+      "learning_rate": 3.0299999999999998e-05,
+      "loss": 7.8844,
       "step": 105
     },
     {
       "epoch": 0.1696,
-      "grad_norm": 20.537715911865234,
-      "learning_rate": 4.08e-05,
-      "loss": 5.4047,
+      "grad_norm": 36.03730010986328,
+      "learning_rate": 3.06e-05,
+      "loss": 7.8807,
       "step": 106
     },
     {
       "epoch": 0.1712,
-      "grad_norm": 17.539196014404297,
-      "learning_rate": 4.12e-05,
-      "loss": 5.2689,
+      "grad_norm": 33.82999801635742,
+      "learning_rate": 3.09e-05,
+      "loss": 7.5575,
       "step": 107
     },
     {
       "epoch": 0.1728,
-      "grad_norm": 14.2318754196167,
-      "learning_rate": 4.1599999999999995e-05,
-      "loss": 4.9925,
+      "grad_norm": 30.55006217956543,
+      "learning_rate": 3.119999999999999e-05,
+      "loss": 7.0033,
       "step": 108
     },
     {
       "epoch": 0.1744,
-      "grad_norm": 13.268582344055176,
-      "learning_rate": 4.2e-05,
-      "loss": 4.9523,
+      "grad_norm": 31.026508331298828,
+      "learning_rate": 3.149999999999999e-05,
+      "loss": 6.9573,
       "step": 109
     },
     {
       "epoch": 0.176,
-      "grad_norm": 11.914827346801758,
-      "learning_rate": 4.24e-05,
-      "loss": 4.8187,
+      "grad_norm": 29.27751922607422,
+      "learning_rate": 3.1799999999999994e-05,
+      "loss": 6.6252,
       "step": 110
     },
     {
       "epoch": 0.1776,
-      "grad_norm": 9.080737113952637,
-      "learning_rate": 4.28e-05,
-      "loss": 4.7312,
+      "grad_norm": 27.643978118896484,
+      "learning_rate": 3.2099999999999994e-05,
+      "loss": 6.4041,
       "step": 111
     },
     {
       "epoch": 0.1792,
-      "grad_norm": 7.438036918640137,
-      "learning_rate": 4.319999999999999e-05,
-      "loss": 4.655,
+      "grad_norm": 25.53025245666504,
+      "learning_rate": 3.2399999999999995e-05,
+      "loss": 6.1614,
       "step": 112
     },
     {
       "epoch": 0.1808,
-      "grad_norm": 7.071867942810059,
-      "learning_rate": 4.3599999999999996e-05,
-      "loss": 4.6643,
+      "grad_norm": 26.29717254638672,
+      "learning_rate": 3.2699999999999995e-05,
+      "loss": 6.1687,
       "step": 113
     },
     {
       "epoch": 0.1824,
-      "grad_norm": 4.7864603996276855,
-      "learning_rate": 4.4e-05,
-      "loss": 4.5668,
+      "grad_norm": 22.344379425048828,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 5.8028,
       "step": 114
     },
     {
       "epoch": 0.184,
-      "grad_norm": 3.781824827194214,
-      "learning_rate": 4.4399999999999995e-05,
-      "loss": 4.503,
+      "grad_norm": 20.374876022338867,
+      "learning_rate": 3.3299999999999996e-05,
+      "loss": 5.5785,
       "step": 115
     },
     {
       "epoch": 0.1856,
-      "grad_norm": 3.6288676261901855,
-      "learning_rate": 4.48e-05,
-      "loss": 4.4637,
+      "grad_norm": 19.226865768432617,
+      "learning_rate": 3.36e-05,
+      "loss": 5.4582,
       "step": 116
     },
     {
       "epoch": 0.1872,
-      "grad_norm": 3.490495443344116,
-      "learning_rate": 4.5199999999999994e-05,
-      "loss": 4.467,
+      "grad_norm": 19.295761108398438,
+      "learning_rate": 3.39e-05,
+      "loss": 5.4439,
       "step": 117
     },
     {
       "epoch": 0.1888,
-      "grad_norm": 3.827871561050415,
-      "learning_rate": 4.56e-05,
-      "loss": 4.4434,
+      "grad_norm": 16.88180923461914,
+      "learning_rate": 3.42e-05,
+      "loss": 5.291,
       "step": 118
     },
     {
       "epoch": 0.1904,
-      "grad_norm": 3.4929349422454834,
-      "learning_rate": 4.599999999999999e-05,
-      "loss": 4.4377,
+      "grad_norm": 18.576812744140625,
+      "learning_rate": 3.45e-05,
+      "loss": 5.3835,
       "step": 119
     },
     {
       "epoch": 0.192,
-      "grad_norm": 3.696927309036255,
-      "learning_rate": 4.6399999999999996e-05,
-      "loss": 4.3986,
+      "grad_norm": 17.378028869628906,
+      "learning_rate": 3.48e-05,
+      "loss": 5.2777,
       "step": 120
     },
     {
       "epoch": 0.1936,
-      "grad_norm": 3.202934980392456,
-      "learning_rate": 4.68e-05,
-      "loss": 4.3563,
+      "grad_norm": 16.458574295043945,
+      "learning_rate": 3.51e-05,
+      "loss": 5.1775,
       "step": 121
     },
     {
       "epoch": 0.1952,
-      "grad_norm": 4.273110866546631,
-      "learning_rate": 4.7199999999999995e-05,
-      "loss": 4.2898,
+      "grad_norm": 11.267019271850586,
+      "learning_rate": 3.539999999999999e-05,
+      "loss": 4.9021,
       "step": 122
     },
     {
       "epoch": 0.1968,
-      "grad_norm": 4.1607346534729,
-      "learning_rate": 4.76e-05,
-      "loss": 4.2119,
+      "grad_norm": 10.104020118713379,
+      "learning_rate": 3.5699999999999994e-05,
+      "loss": 4.812,
       "step": 123
     },
     {
       "epoch": 0.1984,
-      "grad_norm": 4.52202033996582,
-      "learning_rate": 4.7999999999999994e-05,
-      "loss": 4.225,
+      "grad_norm": 7.518466949462891,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 4.7303,
       "step": 124
     },
     {
       "epoch": 0.2,
-      "grad_norm": 4.284320831298828,
-      "learning_rate": 4.84e-05,
-      "loss": 4.1765,
+      "grad_norm": 6.941814422607422,
+      "learning_rate": 3.6299999999999995e-05,
+      "loss": 4.6951,
       "step": 125
     },
     {
       "epoch": 0.2016,
-      "grad_norm": 2.703249454498291,
-      "learning_rate": 4.8799999999999994e-05,
-      "loss": 4.1354,
+      "grad_norm": 8.405250549316406,
+      "learning_rate": 3.6599999999999995e-05,
+      "loss": 4.7108,
       "step": 126
     },
     {
       "epoch": 0.2032,
-      "grad_norm": 2.5615289211273193,
-      "learning_rate": 4.9199999999999997e-05,
-      "loss": 4.1119,
+      "grad_norm": 5.605133056640625,
+      "learning_rate": 3.6899999999999996e-05,
+      "loss": 4.6324,
       "step": 127
     },
     {
       "epoch": 0.2048,
-      "grad_norm": 2.4537241458892822,
-      "learning_rate": 4.959999999999999e-05,
-      "loss": 4.0635,
+      "grad_norm": 4.763094902038574,
+      "learning_rate": 3.7199999999999996e-05,
+      "loss": 4.5848,
       "step": 128
     },
     {
       "epoch": 0.2064,
-      "grad_norm": 2.5134150981903076,
-      "learning_rate": 4.9999999999999996e-05,
-      "loss": 4.0521,
+      "grad_norm": 4.2179646492004395,
+      "learning_rate": 3.75e-05,
+      "loss": 4.5576,
       "step": 129
     },
     {
       "epoch": 0.208,
-      "grad_norm": 3.711264133453369,
-      "learning_rate": 5.04e-05,
-      "loss": 4.0572,
+      "grad_norm": 4.336385726928711,
+      "learning_rate": 3.78e-05,
+      "loss": 4.535,
       "step": 130
     },
     {
       "epoch": 0.2096,
-      "grad_norm": 3.0089399814605713,
-      "learning_rate": 5.0799999999999995e-05,
-      "loss": 4.0094,
+      "grad_norm": 3.6008989810943604,
+      "learning_rate": 3.81e-05,
+      "loss": 4.481,
       "step": 131
     },
     {
       "epoch": 0.2112,
-      "grad_norm": 4.526411533355713,
-      "learning_rate": 5.119999999999999e-05,
-      "loss": 4.0729,
+      "grad_norm": 3.8394198417663574,
+      "learning_rate": 3.84e-05,
+      "loss": 4.5259,
       "step": 132
     },
     {
       "epoch": 0.2128,
-      "grad_norm": 3.459010362625122,
-      "learning_rate": 5.1599999999999994e-05,
-      "loss": 4.0024,
+      "grad_norm": 3.363095998764038,
+      "learning_rate": 3.87e-05,
+      "loss": 4.4499,
       "step": 133
     },
     {
       "epoch": 0.2144,
-      "grad_norm": 4.967794418334961,
-      "learning_rate": 5.2e-05,
-      "loss": 4.0761,
+      "grad_norm": 3.6167714595794678,
+      "learning_rate": 3.9e-05,
+      "loss": 4.4849,
       "step": 134
     },
     {
       "epoch": 0.216,
-      "grad_norm": 1.4037067890167236,
-      "learning_rate": 5.239999999999999e-05,
-      "loss": 3.9801,
+      "grad_norm": 4.084835529327393,
+      "learning_rate": 3.93e-05,
+      "loss": 4.4275,
       "step": 135
     },
     {
       "epoch": 0.2176,
-      "grad_norm": 6.1755781173706055,
-      "learning_rate": 5.279999999999999e-05,
-      "loss": 3.9841,
+      "grad_norm": 3.523148775100708,
+      "learning_rate": 3.96e-05,
+      "loss": 4.3896,
       "step": 136
     },
     {
       "epoch": 0.2192,
-      "grad_norm": 2.2986528873443604,
-      "learning_rate": 5.32e-05,
-      "loss": 4.0046,
+      "grad_norm": 3.151064157485962,
+      "learning_rate": 3.99e-05,
+      "loss": 4.3881,
       "step": 137
     },
     {
       "epoch": 0.2208,
-      "grad_norm": 2.812514066696167,
-      "learning_rate": 5.3599999999999995e-05,
-      "loss": 3.993,
+      "grad_norm": 3.005023717880249,
+      "learning_rate": 4.02e-05,
+      "loss": 4.3544,
       "step": 138
     },
     {
       "epoch": 0.2224,
-      "grad_norm": 1.9895284175872803,
-      "learning_rate": 5.399999999999999e-05,
-      "loss": 4.0296,
+      "grad_norm": 2.9251019954681396,
+      "learning_rate": 4.05e-05,
+      "loss": 4.3701,
       "step": 139
     },
     {
       "epoch": 0.224,
-      "grad_norm": 3.297950506210327,
-      "learning_rate": 5.44e-05,
-      "loss": 3.936,
+      "grad_norm": 4.313841819763184,
+      "learning_rate": 4.08e-05,
+      "loss": 4.2844,
       "step": 140
     },
     {
       "epoch": 0.2256,
-      "grad_norm": 2.8747730255126953,
-      "learning_rate": 5.48e-05,
-      "loss": 3.9037,
+      "grad_norm": 3.625789165496826,
+      "learning_rate": 4.11e-05,
+      "loss": 4.2265,
       "step": 141
     },
     {
       "epoch": 0.2272,
-      "grad_norm": 0.9972714781761169,
-      "learning_rate": 5.519999999999999e-05,
-      "loss": 3.9606,
+      "grad_norm": 2.4950482845306396,
+      "learning_rate": 4.14e-05,
+      "loss": 4.2455,
       "step": 142
     },
     {
       "epoch": 0.2288,
-      "grad_norm": 3.7870514392852783,
-      "learning_rate": 5.559999999999999e-05,
-      "loss": 3.878,
+      "grad_norm": 2.7788918018341064,
+      "learning_rate": 4.17e-05,
+      "loss": 4.1592,
       "step": 143
     },
     {
       "epoch": 0.2304,
-      "grad_norm": 2.812012195587158,
-      "learning_rate": 5.6e-05,
-      "loss": 3.8778,
+      "grad_norm": 2.6029088497161865,
+      "learning_rate": 4.2e-05,
+      "loss": 4.1254,
       "step": 144
     },
     {
       "epoch": 0.232,
-      "grad_norm": 3.2882652282714844,
-      "learning_rate": 5.6399999999999995e-05,
-      "loss": 3.7931,
+      "grad_norm": 2.919585704803467,
+      "learning_rate": 4.229999999999999e-05,
+      "loss": 4.0502,
       "step": 145
     },
     {
       "epoch": 0.2336,
-      "grad_norm": 1.2659672498703003,
-      "learning_rate": 5.679999999999999e-05,
-      "loss": 3.8827,
+      "grad_norm": 2.0354297161102295,
+      "learning_rate": 4.259999999999999e-05,
+      "loss": 4.1037,
       "step": 146
     },
     {
       "epoch": 0.2352,
-      "grad_norm": 3.4999775886535645,
-      "learning_rate": 5.72e-05,
-      "loss": 3.8191,
+      "grad_norm": 2.800039529800415,
+      "learning_rate": 4.289999999999999e-05,
+      "loss": 4.0396,
       "step": 147
     },
     {
       "epoch": 0.2368,
-      "grad_norm": 1.5007144212722778,
-      "learning_rate": 5.76e-05,
-      "loss": 3.8964,
+      "grad_norm": 3.689743995666504,
+      "learning_rate": 4.319999999999999e-05,
+      "loss": 4.1239,
       "step": 148
     },
     {
       "epoch": 0.2384,
-      "grad_norm": 1.7462321519851685,
-      "learning_rate": 5.7999999999999994e-05,
-      "loss": 3.9211,
+      "grad_norm": 3.157559871673584,
+      "learning_rate": 4.3499999999999993e-05,
+      "loss": 4.1312,
       "step": 149
     },
     {
       "epoch": 0.24,
-      "grad_norm": 1.0943163633346558,
-      "learning_rate": 5.839999999999999e-05,
-      "loss": 3.9544,
+      "grad_norm": 1.6889827251434326,
+      "learning_rate": 4.3799999999999994e-05,
+      "loss": 4.1446,
       "step": 150
     },
     {
       "epoch": 0.2416,
-      "grad_norm": 49.87346649169922,
-      "learning_rate": 5.88e-05,
-      "loss": 5.3277,
+      "grad_norm": 43.56467819213867,
+      "learning_rate": 4.4099999999999995e-05,
+      "loss": 5.576,
       "step": 151
     },
     {
       "epoch": 0.2432,
-      "grad_norm": 19.33259391784668,
-      "learning_rate": 5.9199999999999996e-05,
-      "loss": 4.1933,
+      "grad_norm": 17.566373825073242,
+      "learning_rate": 4.4399999999999995e-05,
+      "loss": 4.3442,
       "step": 152
     },
     {
       "epoch": 0.2448,
-      "grad_norm": 42.26847839355469,
-      "learning_rate": 5.959999999999999e-05,
-      "loss": 4.3403,
+      "grad_norm": 19.273839950561523,
+      "learning_rate": 4.4699999999999996e-05,
+      "loss": 4.4241,
       "step": 153
     },
     {
       "epoch": 0.2464,
-      "grad_norm": 12.728076934814453,
-      "learning_rate": 5.9999999999999995e-05,
-      "loss": 3.996,
+      "grad_norm": 43.27154541015625,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 4.0996,
       "step": 154
     },
     {
       "epoch": 0.248,
-      "grad_norm": 26.57826805114746,
-      "learning_rate": 6.04e-05,
-      "loss": 4.5252,
+      "grad_norm": 22.586294174194336,
+      "learning_rate": 4.5299999999999997e-05,
+      "loss": 4.5675,
       "step": 155
     },
     {
       "epoch": 0.2496,
-      "grad_norm": 7.899193286895752,
-      "learning_rate": 6.0799999999999994e-05,
-      "loss": 3.8146,
+      "grad_norm": 5.962088108062744,
+      "learning_rate": 4.56e-05,
+      "loss": 3.9192,
       "step": 156
     },
     {
       "epoch": 0.2512,
-      "grad_norm": 7.121301651000977,
-      "learning_rate": 6.12e-05,
-      "loss": 3.8425,
+      "grad_norm": 5.2486701011657715,
+      "learning_rate": 4.59e-05,
+      "loss": 3.929,
       "step": 157
     },
     {
       "epoch": 0.2528,
-      "grad_norm": 17.966394424438477,
-      "learning_rate": 6.159999999999999e-05,
-      "loss": 3.8739,
+      "grad_norm": 6.9419331550598145,
+      "learning_rate": 4.62e-05,
+      "loss": 3.9572,
       "step": 158
     },
     {
       "epoch": 0.2544,
-      "grad_norm": 3.4284541606903076,
-      "learning_rate": 6.199999999999999e-05,
-      "loss": 3.817,
+      "grad_norm": 10.546914100646973,
+      "learning_rate": 4.65e-05,
+      "loss": 3.9106,
       "step": 159
     },
     {
       "epoch": 0.256,
-      "grad_norm": 0.9217383861541748,
-      "learning_rate": 6.239999999999999e-05,
-      "loss": 3.7066,
+      "grad_norm": 1.4478720426559448,
+      "learning_rate": 4.68e-05,
+      "loss": 3.816,
       "step": 160
     },
     {
       "epoch": 0.2576,
-      "grad_norm": 1.034417986869812,
-      "learning_rate": 6.28e-05,
-      "loss": 3.7111,
+      "grad_norm": 1.5371733903884888,
+      "learning_rate": 4.709999999999999e-05,
+      "loss": 3.8152,
       "step": 161
     },
     {
       "epoch": 0.2592,
-      "grad_norm": 1.0513995885849,
-      "learning_rate": 6.319999999999999e-05,
-      "loss": 3.7579,
+      "grad_norm": 1.1498860120773315,
+      "learning_rate": 4.7399999999999993e-05,
+      "loss": 3.8535,
       "step": 162
     },
     {
       "epoch": 0.2608,
-      "grad_norm": 3.502819776535034,
-      "learning_rate": 6.359999999999999e-05,
-      "loss": 3.7508,
+      "grad_norm": 3.339113473892212,
+      "learning_rate": 4.7699999999999994e-05,
+      "loss": 3.8388,
       "step": 163
     },
     {
       "epoch": 0.2624,
-      "grad_norm": 2.819303035736084,
-      "learning_rate": 6.4e-05,
-      "loss": 3.697,
+      "grad_norm": 2.404588460922241,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 3.7853,
       "step": 164
     },
     {
       "epoch": 0.264,
-      "grad_norm": 2.330233097076416,
-      "learning_rate": 6.44e-05,
-      "loss": 3.7655,
+      "grad_norm": 1.7241195440292358,
+      "learning_rate": 4.8299999999999995e-05,
+      "loss": 3.8405,
       "step": 165
     },
     {
       "epoch": 0.2656,
-      "grad_norm": 1.2856422662734985,
-      "learning_rate": 6.479999999999999e-05,
-      "loss": 3.7024,
+      "grad_norm": 1.0453760623931885,
+      "learning_rate": 4.8599999999999995e-05,
+      "loss": 3.7872,
       "step": 166
     },
     {
       "epoch": 0.2672,
-      "grad_norm": 0.9377660155296326,
-      "learning_rate": 6.519999999999999e-05,
-      "loss": 3.7571,
+      "grad_norm": 0.8506628274917603,
+      "learning_rate": 4.8899999999999996e-05,
+      "loss": 3.8375,
       "step": 167
     },
     {
       "epoch": 0.2688,
-      "grad_norm": 2.6969499588012695,
-      "learning_rate": 6.56e-05,
-      "loss": 3.6941,
+      "grad_norm": 2.324955940246582,
+      "learning_rate": 4.9199999999999997e-05,
+      "loss": 3.7611,
       "step": 168
     },
     {
       "epoch": 0.2704,
-      "grad_norm": 3.896618604660034,
-      "learning_rate": 6.599999999999999e-05,
-      "loss": 3.7757,
+      "grad_norm": 3.906370162963867,
+      "learning_rate": 4.95e-05,
+      "loss": 3.8567,
       "step": 169
     },
     {
       "epoch": 0.272,
-      "grad_norm": 0.6105890870094299,
-      "learning_rate": 6.639999999999999e-05,
-      "loss": 3.7015,
+      "grad_norm": 0.7771504521369934,
+      "learning_rate": 4.98e-05,
+      "loss": 3.7718,
       "step": 170
     },
     {
       "epoch": 0.2736,
-      "grad_norm": 1.8918788433074951,
-      "learning_rate": 6.68e-05,
-      "loss": 3.6959,
+      "grad_norm": 1.1586111783981323,
+      "learning_rate": 5.01e-05,
+      "loss": 3.7604,
       "step": 171
     },
     {
       "epoch": 0.2752,
-      "grad_norm": 1.5491015911102295,
-      "learning_rate": 6.72e-05,
-      "loss": 3.6853,
+      "grad_norm": 1.0725429058074951,
+      "learning_rate": 5.04e-05,
+      "loss": 3.7399,
       "step": 172
     },
     {
       "epoch": 0.2768,
-      "grad_norm": 2.8067069053649902,
-      "learning_rate": 6.759999999999999e-05,
-      "loss": 3.7012,
+      "grad_norm": 2.3980965614318848,
+      "learning_rate": 5.07e-05,
+      "loss": 3.7637,
       "step": 173
     },
     {
       "epoch": 0.2784,
-      "grad_norm": 17.379011154174805,
-      "learning_rate": 6.799999999999999e-05,
-      "loss": 3.7106,
+      "grad_norm": 1.2434422969818115,
+      "learning_rate": 5.1e-05,
+      "loss": 3.7556,
       "step": 174
     },
     {
       "epoch": 0.28,
-      "grad_norm": 2.3803513050079346,
-      "learning_rate": 6.84e-05,
-      "loss": 3.7663,
+      "grad_norm": 2.1702017784118652,
+      "learning_rate": 5.13e-05,
+      "loss": 3.8249,
       "step": 175
     },
     {
       "epoch": 0.2816,
-      "grad_norm": 0.6946349740028381,
-      "learning_rate": 6.879999999999999e-05,
-      "loss": 3.7089,
+      "grad_norm": 0.7948981523513794,
+      "learning_rate": 5.1599999999999994e-05,
+      "loss": 3.7648,
       "step": 176
     },
     {
       "epoch": 0.2832,
-      "grad_norm": 2.5919649600982666,
-      "learning_rate": 6.919999999999999e-05,
-      "loss": 3.6104,
+      "grad_norm": 2.3048489093780518,
+      "learning_rate": 5.1899999999999994e-05,
+      "loss": 3.6571,
       "step": 177
     },
     {
       "epoch": 0.2848,
-      "grad_norm": 1.1172287464141846,
-      "learning_rate": 6.96e-05,
-      "loss": 3.7244,
+      "grad_norm": 0.9848898649215698,
+      "learning_rate": 5.2199999999999995e-05,
+      "loss": 3.7757,
       "step": 178
     },
     {
       "epoch": 0.2864,
-      "grad_norm": 2.474264144897461,
-      "learning_rate": 7e-05,
-      "loss": 3.7115,
+      "grad_norm": 2.904839515686035,
+      "learning_rate": 5.2499999999999995e-05,
+      "loss": 3.7699,
       "step": 179
     },
     {
       "epoch": 0.288,
-      "grad_norm": 2.681856870651245,
-      "learning_rate": 7.039999999999999e-05,
-      "loss": 3.6786,
+      "grad_norm": 1.1673673391342163,
+      "learning_rate": 5.279999999999999e-05,
+      "loss": 3.7191,
       "step": 180
     },
     {
       "epoch": 0.2896,
-      "grad_norm": 1.6489158868789673,
-      "learning_rate": 7.079999999999999e-05,
-      "loss": 3.6661,
+      "grad_norm": 1.269823670387268,
+      "learning_rate": 5.309999999999999e-05,
+      "loss": 3.7021,
       "step": 181
     },
     {
       "epoch": 0.2912,
-      "grad_norm": 1.1805118322372437,
-      "learning_rate": 7.12e-05,
-      "loss": 3.654,
+      "grad_norm": 0.9047891497612,
+      "learning_rate": 5.339999999999999e-05,
+      "loss": 3.7024,
       "step": 182
     },
     {
       "epoch": 0.2928,
-      "grad_norm": 2.3934733867645264,
-      "learning_rate": 7.159999999999999e-05,
-      "loss": 3.7694,
+      "grad_norm": 2.7220137119293213,
+      "learning_rate": 5.369999999999999e-05,
+      "loss": 3.8191,
       "step": 183
     },
     {
       "epoch": 0.2944,
-      "grad_norm": 1.6547563076019287,
-      "learning_rate": 7.199999999999999e-05,
-      "loss": 3.6795,
+      "grad_norm": 1.4135470390319824,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 3.7193,
       "step": 184
     },
     {
       "epoch": 0.296,
-      "grad_norm": 3.933615207672119,
-      "learning_rate": 7.24e-05,
-      "loss": 3.7822,
+      "grad_norm": 4.183174133300781,
+      "learning_rate": 5.429999999999999e-05,
+      "loss": 3.832,
       "step": 185
     },
     {
       "epoch": 0.2976,
-      "grad_norm": 2.2447006702423096,
-      "learning_rate": 7.28e-05,
-      "loss": 3.7498,
+      "grad_norm": 2.3090178966522217,
+      "learning_rate": 5.459999999999999e-05,
+      "loss": 3.7992,
       "step": 186
     },
     {
       "epoch": 0.2992,
-      "grad_norm": 1.272269606590271,
-      "learning_rate": 7.319999999999999e-05,
-      "loss": 3.7266,
+      "grad_norm": 1.39043128490448,
+      "learning_rate": 5.489999999999999e-05,
+      "loss": 3.7616,
       "step": 187
     },
     {
       "epoch": 0.3008,
-      "grad_norm": 0.6136051416397095,
-      "learning_rate": 7.359999999999999e-05,
-      "loss": 3.6942,
+      "grad_norm": 0.6333518028259277,
+      "learning_rate": 5.519999999999999e-05,
+      "loss": 3.7379,
       "step": 188
     },
     {
       "epoch": 0.3024,
-      "grad_norm": 0.6757500767707825,
-      "learning_rate": 7.4e-05,
-      "loss": 3.6945,
+      "grad_norm": 0.6841683387756348,
+      "learning_rate": 5.5499999999999994e-05,
+      "loss": 3.7347,
       "step": 189
     },
     {
       "epoch": 0.304,
-      "grad_norm": 1.0322827100753784,
-      "learning_rate": 7.439999999999999e-05,
-      "loss": 3.6552,
+      "grad_norm": 0.857160747051239,
+      "learning_rate": 5.5799999999999994e-05,
+      "loss": 3.6922,
       "step": 190
     },
     {
       "epoch": 0.3056,
-      "grad_norm": 0.9092890620231628,
-      "learning_rate": 7.479999999999999e-05,
-      "loss": 3.7145,
+      "grad_norm": 1.119098424911499,
+      "learning_rate": 5.6099999999999995e-05,
+      "loss": 3.7586,
       "step": 191
     },
     {
       "epoch": 0.3072,
-      "grad_norm": 0.8890368342399597,
-      "learning_rate": 7.519999999999998e-05,
-      "loss": 3.765,
+      "grad_norm": 0.9412086606025696,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 3.8095,
       "step": 192
     },
     {
       "epoch": 0.3088,
-      "grad_norm": 1.3483480215072632,
-      "learning_rate": 7.56e-05,
-      "loss": 3.771,
+      "grad_norm": 0.7174120545387268,
+      "learning_rate": 5.6699999999999996e-05,
+      "loss": 3.8177,
       "step": 193
     },
     {
       "epoch": 0.3104,
-      "grad_norm": 3.8980510234832764,
-      "learning_rate": 7.6e-05,
-      "loss": 3.7792,
+      "grad_norm": 3.7620341777801514,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 3.8253,
       "step": 194
     },
     {
       "epoch": 0.312,
-      "grad_norm": 0.7139620780944824,
-      "learning_rate": 7.639999999999999e-05,
-      "loss": 3.7156,
+      "grad_norm": 0.6953389644622803,
+      "learning_rate": 5.73e-05,
+      "loss": 3.7556,
       "step": 195
     },
     {
       "epoch": 0.3136,
-      "grad_norm": 1.751401424407959,
-      "learning_rate": 7.68e-05,
-      "loss": 3.6977,
+      "grad_norm": 2.053453207015991,
+      "learning_rate": 5.76e-05,
+      "loss": 3.7528,
       "step": 196
     },
     {
       "epoch": 0.3152,
-      "grad_norm": 1.1341018676757812,
-      "learning_rate": 7.72e-05,
-      "loss": 3.6716,
+      "grad_norm": 0.926834225654602,
+      "learning_rate": 5.79e-05,
+      "loss": 3.7177,
       "step": 197
     },
     {
       "epoch": 0.3168,
-      "grad_norm": 1.0171042680740356,
-      "learning_rate": 7.759999999999999e-05,
-      "loss": 3.719,
+      "grad_norm": 0.9621994495391846,
+      "learning_rate": 5.82e-05,
+      "loss": 3.7611,
       "step": 198
     },
     {
       "epoch": 0.3184,
-      "grad_norm": 2.265624523162842,
-      "learning_rate": 7.8e-05,
-      "loss": 3.7401,
+      "grad_norm": 2.589616537094116,
+      "learning_rate": 5.85e-05,
+      "loss": 3.7764,
       "step": 199
     },
     {
       "epoch": 0.32,
       "grad_norm": NaN,
-      "learning_rate": 7.8e-05,
-      "loss": 4.4471,
-      "step": 200
-    },
-    {
-      "epoch": 0.32,
-      "eval_cer": 0.9811814884202974,
-      "eval_loss": 3.759681463241577,
-      "eval_runtime": 157.878,
-      "eval_samples_per_second": 19.863,
-      "eval_steps_per_second": 1.241,
-      "eval_wer": 1.0,
+      "learning_rate": 5.85e-05,
+      "loss": 4.5885,
       "step": 200
     },
     {
       "epoch": 0.3216,
-      "grad_norm": 40.01380920410156,
-      "learning_rate": 7.839999999999998e-05,
-      "loss": 4.8451,
+      "grad_norm": 36.955745697021484,
+      "learning_rate": 5.88e-05,
+      "loss": 4.8317,
       "step": 201
     },
     {
       "epoch": 0.3232,
-      "grad_norm": 16.052330017089844,
-      "learning_rate": 7.879999999999999e-05,
-      "loss": 3.9661,
+      "grad_norm": 14.525715827941895,
+      "learning_rate": 5.91e-05,
+      "loss": 3.9233,
       "step": 202
     },
     {
       "epoch": 0.3248,
-      "grad_norm": 31.00684928894043,
-      "learning_rate": 7.92e-05,
-      "loss": 4.454,
+      "grad_norm": 26.20575714111328,
+      "learning_rate": 5.94e-05,
+      "loss": 4.3815,
       "step": 203
     },
     {
       "epoch": 0.3264,
-      "grad_norm": 6.007049083709717,
-      "learning_rate": 7.959999999999998e-05,
-      "loss": 3.7405,
+      "grad_norm": 6.467894554138184,
+      "learning_rate": 5.97e-05,
+      "loss": 3.768,
       "step": 204
     },
     {
       "epoch": 0.328,
-      "grad_norm": 8.176580429077148,
-      "learning_rate": 7.999999999999999e-05,
-      "loss": 3.7768,
+      "grad_norm": 6.852605819702148,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 3.7986,
       "step": 205
     },
     {
       "epoch": 0.3296,
-      "grad_norm": 10.413802146911621,
-      "learning_rate": 8.04e-05,
-      "loss": 3.9944,
+      "grad_norm": 10.890352249145508,
+      "learning_rate": 6.0299999999999995e-05,
+      "loss": 4.0176,
       "step": 206
     },
     {
       "epoch": 0.3312,
-      "grad_norm": 3.2589361667633057,
-      "learning_rate": 8.079999999999999e-05,
-      "loss": 3.7885,
+      "grad_norm": 5.478235721588135,
+      "learning_rate": 6.0599999999999996e-05,
+      "loss": 3.8291,
       "step": 207
     },
     {
       "epoch": 0.3328,
-      "grad_norm": 4.633168697357178,
-      "learning_rate": 8.12e-05,
-      "loss": 3.7046,
+      "grad_norm": 2.3643927574157715,
+      "learning_rate": 6.0899999999999996e-05,
+      "loss": 3.692,
       "step": 208
     },
     {
       "epoch": 0.3344,
-      "grad_norm": 3.281346321105957,
-      "learning_rate": 8.16e-05,
-      "loss": 3.8043,
+      "grad_norm": 0.9052429795265198,
+      "learning_rate": 6.12e-05,
+      "loss": 3.8016,
       "step": 209
     },
     {
       "epoch": 0.336,
-      "grad_norm": 4.552253246307373,
-      "learning_rate": 8.199999999999999e-05,
-      "loss": 3.7858,
+      "grad_norm": 2.090831995010376,
+      "learning_rate": 6.149999999999999e-05,
+      "loss": 3.7758,
       "step": 210
     },
     {
       "epoch": 0.3376,
-      "grad_norm": 2.2564077377319336,
-      "learning_rate": 8.24e-05,
-      "loss": 3.8172,
+      "grad_norm": 1.3076395988464355,
+      "learning_rate": 6.18e-05,
+      "loss": 3.8275,
       "step": 211
     },
     {
       "epoch": 0.3392,
-      "grad_norm": 4.931669235229492,
-      "learning_rate": 8.28e-05,
-      "loss": 3.7392,
+      "grad_norm": 3.603741407394409,
+      "learning_rate": 6.209999999999999e-05,
+      "loss": 3.7251,
       "step": 212
     },
     {
       "epoch": 0.3408,
-      "grad_norm": 5.935787677764893,
-      "learning_rate": 8.319999999999999e-05,
-      "loss": 3.6674,
+      "grad_norm": 5.31221866607666,
+      "learning_rate": 6.239999999999999e-05,
+      "loss": 3.6644,
       "step": 213
     },
     {
       "epoch": 0.3424,
-      "grad_norm": 5.499577045440674,
-      "learning_rate": 8.36e-05,
-      "loss": 3.6903,
+      "grad_norm": 5.733158588409424,
+      "learning_rate": 6.269999999999999e-05,
+      "loss": 3.7033,
       "step": 214
     },
     {
       "epoch": 0.344,
-      "grad_norm": 2.0088694095611572,
-      "learning_rate": 8.4e-05,
-      "loss": 3.6526,
+      "grad_norm": 3.1695661544799805,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 3.6704,
       "step": 215
     },
     {
       "epoch": 0.3456,
-      "grad_norm": 1.7222843170166016,
-      "learning_rate": 8.439999999999999e-05,
-      "loss": 3.6685,
+      "grad_norm": 1.594991683959961,
+      "learning_rate": 6.33e-05,
+      "loss": 3.6788,
       "step": 216
     },
     {
       "epoch": 0.3472,
-      "grad_norm": 1.0298047065734863,
-      "learning_rate": 8.48e-05,
-      "loss": 3.6548,
+      "grad_norm": 1.6268411874771118,
+      "learning_rate": 6.359999999999999e-05,
+      "loss": 3.6691,
       "step": 217
     },
     {
       "epoch": 0.3488,
-      "grad_norm": 0.6882361173629761,
-      "learning_rate": 8.519999999999998e-05,
-      "loss": 3.5886,
+      "grad_norm": 1.526171326637268,
+      "learning_rate": 6.39e-05,
+      "loss": 3.613,
       "step": 218
     },
     {
       "epoch": 0.3504,
-      "grad_norm": 3.674741744995117,
-      "learning_rate": 8.56e-05,
-      "loss": 3.6667,
+      "grad_norm": 4.093273162841797,
+      "learning_rate": 6.419999999999999e-05,
+      "loss": 3.6907,
       "step": 219
     },
     {
       "epoch": 0.352,
-      "grad_norm": 0.8755781054496765,
-      "learning_rate": 8.6e-05,
-      "loss": 3.5641,
+      "grad_norm": 2.1383774280548096,
+      "learning_rate": 6.45e-05,
+      "loss": 3.5904,
       "step": 220
     },
     {
       "epoch": 0.3536,
-      "grad_norm": 1.6391894817352295,
-      "learning_rate": 8.639999999999999e-05,
-      "loss": 3.598,
+      "grad_norm": 4.359639644622803,
+      "learning_rate": 6.479999999999999e-05,
+      "loss": 3.6432,
       "step": 221
     },
     {
       "epoch": 0.3552,
-      "grad_norm": 4.3577165603637695,
-      "learning_rate": 8.68e-05,
-      "loss": 3.7042,
+      "grad_norm": 7.548089027404785,
+      "learning_rate": 6.51e-05,
+      "loss": 3.7759,
       "step": 222
     },
     {
       "epoch": 0.3568,
-      "grad_norm": 7.5377960205078125,
-      "learning_rate": 8.719999999999999e-05,
-      "loss": 3.7538,
+      "grad_norm": 10.02462100982666,
+      "learning_rate": 6.539999999999999e-05,
+      "loss": 3.8324,
       "step": 223
     },
     {
       "epoch": 0.3584,
-      "grad_norm": 0.9260856509208679,
-      "learning_rate": 8.759999999999999e-05,
-      "loss": 3.5839,
+      "grad_norm": 2.3231656551361084,
+      "learning_rate": 6.57e-05,
+      "loss": 3.6139,
       "step": 224
     },
     {
       "epoch": 0.36,
-      "grad_norm": 1.126509189605713,
-      "learning_rate": 8.8e-05,
-      "loss": 3.7167,
+      "grad_norm": 2.966249942779541,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 3.743,
       "step": 225
     },
     {
       "epoch": 0.3616,
-      "grad_norm": 0.5898628830909729,
-      "learning_rate": 8.84e-05,
-      "loss": 3.6411,
+      "grad_norm": 1.6161023378372192,
+      "learning_rate": 6.63e-05,
+      "loss": 3.6788,
       "step": 226
     },
     {
       "epoch": 0.3632,
-      "grad_norm": 1.984059453010559,
-      "learning_rate": 8.879999999999999e-05,
-      "loss": 3.8212,
+      "grad_norm": 2.224886655807495,
+      "learning_rate": 6.659999999999999e-05,
+      "loss": 3.8758,
       "step": 227
     },
     {
       "epoch": 0.3648,
-      "grad_norm": 2.239217519760132,
-      "learning_rate": 8.92e-05,
-      "loss": 3.6481,
+      "grad_norm": 3.000509262084961,
+      "learning_rate": 6.69e-05,
+      "loss": 3.6772,
       "step": 228
     },
     {
       "epoch": 0.3664,
-      "grad_norm": 0.8924047946929932,
-      "learning_rate": 8.96e-05,
-      "loss": 3.6647,
+      "grad_norm": 2.5037214756011963,
+      "learning_rate": 6.72e-05,
+      "loss": 3.7001,
       "step": 229
     },
     {
       "epoch": 0.368,
-      "grad_norm": 2.23852801322937,
-      "learning_rate": 8.999999999999999e-05,
-      "loss": 3.605,
+      "grad_norm": 4.560110092163086,
+      "learning_rate": 6.75e-05,
+      "loss": 3.6615,
       "step": 230
     },
     {
       "epoch": 0.3696,
-      "grad_norm": 1.5995547771453857,
-      "learning_rate": 9.039999999999999e-05,
-      "loss": 3.6514,
+      "grad_norm": 5.233952045440674,
+      "learning_rate": 6.78e-05,
+      "loss": 3.7162,
       "step": 231
     },
     {
       "epoch": 0.3712,
-      "grad_norm": 1.6624646186828613,
-      "learning_rate": 9.079999999999998e-05,
-      "loss": 3.5983,
+      "grad_norm": 4.030121803283691,
+      "learning_rate": 6.81e-05,
+      "loss": 3.6467,
       "step": 232
     },
     {
       "epoch": 0.3728,
-      "grad_norm": 2.7629284858703613,
-      "learning_rate": 9.12e-05,
-      "loss": 3.6708,
+      "grad_norm": 3.1937036514282227,
+      "learning_rate": 6.84e-05,
+      "loss": 3.693,
       "step": 233
     },
     {
       "epoch": 0.3744,
-      "grad_norm": 5.317010879516602,
-      "learning_rate": 9.159999999999999e-05,
-      "loss": 3.8436,
+      "grad_norm": 1.4005098342895508,
+      "learning_rate": 6.87e-05,
+      "loss": 3.8877,
       "step": 234
     },
     {
       "epoch": 0.376,
-      "grad_norm": 1.7832276821136475,
-      "learning_rate": 9.199999999999999e-05,
-      "loss": 3.5452,
+      "grad_norm": 3.893272876739502,
+      "learning_rate": 6.9e-05,
+      "loss": 3.593,
       "step": 235
     },
     {
       "epoch": 0.3776,
-      "grad_norm": 0.7123947143554688,
-      "learning_rate": 9.24e-05,
-      "loss": 3.5936,
+      "grad_norm": 0.6252413988113403,
+      "learning_rate": 6.93e-05,
+      "loss": 3.6299,
       "step": 236
     },
     {
       "epoch": 0.3792,
-      "grad_norm": 1.8740249872207642,
-      "learning_rate": 9.279999999999999e-05,
-      "loss": 3.5139,
+      "grad_norm": 0.9029965400695801,
+      "learning_rate": 6.96e-05,
+      "loss": 3.5456,
       "step": 237
     },
     {
       "epoch": 0.3808,
-      "grad_norm": 0.8917979001998901,
-      "learning_rate": 9.319999999999999e-05,
-      "loss": 3.6048,
+      "grad_norm": 1.729459285736084,
+      "learning_rate": 6.989999999999999e-05,
+      "loss": 3.6351,
       "step": 238
     },
     {
       "epoch": 0.3824,
-      "grad_norm": 1.3209342956542969,
-      "learning_rate": 9.36e-05,
-      "loss": 3.5569,
+      "grad_norm": 4.091501235961914,
+      "learning_rate": 7.02e-05,
+      "loss": 3.6169,
       "step": 239
     },
     {
       "epoch": 0.384,
-      "grad_norm": 2.313504934310913,
-      "learning_rate": 9.4e-05,
-      "loss": 3.6239,
+      "grad_norm": 5.327137470245361,
+      "learning_rate": 7.049999999999999e-05,
+      "loss": 3.695,
       "step": 240
     },
     {
       "epoch": 0.3856,
-      "grad_norm": 2.1547036170959473,
-      "learning_rate": 9.439999999999999e-05,
-      "loss": 3.6724,
+      "grad_norm": 5.34423303604126,
+      "learning_rate": 7.079999999999999e-05,
+      "loss": 3.7424,
       "step": 241
     },
     {
       "epoch": 0.3872,
-      "grad_norm": 0.8160881400108337,
-      "learning_rate": 9.479999999999999e-05,
-      "loss": 3.6857,
+      "grad_norm": 3.87858510017395,
+      "learning_rate": 7.11e-05,
+      "loss": 3.7498,
       "step": 242
     },
     {
       "epoch": 0.3888,
-      "grad_norm": 2.1986379623413086,
-      "learning_rate": 9.52e-05,
-      "loss": 3.5633,
+      "grad_norm": 1.0561057329177856,
+      "learning_rate": 7.139999999999999e-05,
+      "loss": 3.6034,
       "step": 243
     },
     {
       "epoch": 0.3904,
-      "grad_norm": 1.8258991241455078,
-      "learning_rate": 9.559999999999999e-05,
-      "loss": 3.6461,
+      "grad_norm": 0.686858594417572,
+      "learning_rate": 7.17e-05,
+      "loss": 3.6911,
       "step": 244
     },
     {
       "epoch": 0.392,
-      "grad_norm": 1.290541410446167,
-      "learning_rate": 9.599999999999999e-05,
-      "loss": 3.7378,
+      "grad_norm": 1.0606696605682373,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 3.7628,
       "step": 245
     },
     {
       "epoch": 0.3936,
-      "grad_norm": 2.4464797973632812,
-      "learning_rate": 9.64e-05,
-      "loss": 3.584,
+      "grad_norm": 4.318437576293945,
+      "learning_rate": 7.23e-05,
+      "loss": 3.6438,
       "step": 246
     },
     {
       "epoch": 0.3952,
-      "grad_norm": 1.3107831478118896,
-      "learning_rate": 9.68e-05,
-      "loss": 3.6602,
+      "grad_norm": 3.086888074874878,
+      "learning_rate": 7.259999999999999e-05,
+      "loss": 3.7093,
       "step": 247
     },
     {
       "epoch": 0.3968,
-      "grad_norm": 1.3498066663742065,
-      "learning_rate": 9.719999999999999e-05,
-      "loss": 3.6242,
+      "grad_norm": 4.907326698303223,
+      "learning_rate": 7.29e-05,
+      "loss": 3.709,
       "step": 248
     },
     {
       "epoch": 0.3984,
-      "grad_norm": 1.8473575115203857,
-      "learning_rate": 9.759999999999999e-05,
-      "loss": 3.6759,
+      "grad_norm": 5.898931503295898,
+      "learning_rate": 7.319999999999999e-05,
+      "loss": 3.775,
       "step": 249
     },
     {
       "epoch": 0.4,
-      "grad_norm": 1.866666555404663,
-      "learning_rate": 9.799999999999998e-05,
-      "loss": 3.753,
+      "grad_norm": 6.163908004760742,
+      "learning_rate": 7.35e-05,
+      "loss": 3.8484,
       "step": 250
     },
     {
       "epoch": 0.4016,
-      "grad_norm": 100.26708984375,
-      "learning_rate": 9.839999999999999e-05,
-      "loss": 7.058,
+      "grad_norm": 61.644100189208984,
+      "learning_rate": 7.379999999999999e-05,
+      "loss": 5.9524,
       "step": 251
     },
     {
       "epoch": 0.4032,
-      "grad_norm": 24.822065353393555,
-      "learning_rate": 9.879999999999999e-05,
-      "loss": 4.1968,
+      "grad_norm": 12.312813758850098,
+      "learning_rate": 7.41e-05,
+      "loss": 3.862,
       "step": 252
     },
     {
       "epoch": 0.4048,
-      "grad_norm": 22.42405128479004,
-      "learning_rate": 9.919999999999999e-05,
-      "loss": 4.3078,
+      "grad_norm": 11.08806037902832,
+      "learning_rate": 7.439999999999999e-05,
+      "loss": 4.0092,
       "step": 253
     },
     {
       "epoch": 0.4064,
-      "grad_norm": 45.404808044433594,
-      "learning_rate": 9.96e-05,
-      "loss": 4.1147,
+      "grad_norm": 10.701560974121094,
+      "learning_rate": 7.47e-05,
+      "loss": 3.9019,
       "step": 254
     },
     {
       "epoch": 0.408,
-      "grad_norm": 18.83899688720703,
-      "learning_rate": 9.999999999999999e-05,
-      "loss": 4.0957,
+      "grad_norm": 8.322529792785645,
+      "learning_rate": 7.5e-05,
+      "loss": 3.8767,
       "step": 255
     },
     {
       "epoch": 0.4096,
-      "grad_norm": 25.831239700317383,
-      "learning_rate": 0.00010039999999999999,
-      "loss": 4.1593,
+      "grad_norm": 11.361836433410645,
+      "learning_rate": 7.529999999999999e-05,
+      "loss": 4.0091,
       "step": 256
     },
     {
       "epoch": 0.4112,
-      "grad_norm": 8.648876190185547,
-      "learning_rate": 0.0001008,
-      "loss": 3.7909,
+      "grad_norm": 1.967650055885315,
+      "learning_rate": 7.56e-05,
+      "loss": 3.7382,
       "step": 257
     },
     {
       "epoch": 0.4128,
-      "grad_norm": 8.114570617675781,
-      "learning_rate": 0.0001012,
-      "loss": 3.9005,
+      "grad_norm": 2.5838770866394043,
+      "learning_rate": 7.589999999999999e-05,
+      "loss": 3.8723,
       "step": 258
     },
     {
       "epoch": 0.4144,
-      "grad_norm": 4.483002662658691,
-      "learning_rate": 0.00010159999999999999,
-      "loss": 3.6031,
+      "grad_norm": 7.383574962615967,
+      "learning_rate": 7.62e-05,
+      "loss": 3.6693,
       "step": 259
     },
     {
       "epoch": 0.416,
-      "grad_norm": 1.673669457435608,
-      "learning_rate": 0.000102,
-      "loss": 3.7067,
+      "grad_norm": 3.9820754528045654,
+      "learning_rate": 7.649999999999999e-05,
+      "loss": 3.74,
       "step": 260
     },
     {
       "epoch": 0.4176,
-      "grad_norm": 3.5352141857147217,
-      "learning_rate": 0.00010239999999999998,
-      "loss": 3.7009,
+      "grad_norm": 5.922429084777832,
+      "learning_rate": 7.68e-05,
+      "loss": 3.7551,
       "step": 261
     },
     {
       "epoch": 0.4192,
-      "grad_norm": 7.101142883300781,
-      "learning_rate": 0.00010279999999999999,
-      "loss": 3.6875,
+      "grad_norm": 8.945326805114746,
+      "learning_rate": 7.709999999999999e-05,
+      "loss": 3.759,
       "step": 262
     },
     {
       "epoch": 0.4208,
-      "grad_norm": 1.425631046295166,
-      "learning_rate": 0.00010319999999999999,
-      "loss": 3.7991,
+      "grad_norm": 3.62739896774292,
+      "learning_rate": 7.74e-05,
+      "loss": 3.8278,
       "step": 263
     },
     {
       "epoch": 0.4224,
-      "grad_norm": 3.5846240520477295,
-      "learning_rate": 0.00010359999999999998,
-      "loss": 3.73,
+      "grad_norm": 5.89738130569458,
+      "learning_rate": 7.769999999999999e-05,
+      "loss": 3.7793,
       "step": 264
     },
     {
       "epoch": 0.424,
-      "grad_norm": 4.524421215057373,
-      "learning_rate": 0.000104,
-      "loss": 3.5848,
+      "grad_norm": 6.884239673614502,
+      "learning_rate": 7.8e-05,
+      "loss": 3.654,
       "step": 265
     },
     {
       "epoch": 0.4256,
-      "grad_norm": 2.8252696990966797,
-      "learning_rate": 0.00010439999999999999,
-      "loss": 3.6126,
+      "grad_norm": 5.8110857009887695,
+      "learning_rate": 7.829999999999999e-05,
+      "loss": 3.6802,
       "step": 266
     },
     {
       "epoch": 0.4272,
-      "grad_norm": 1.503335952758789,
-      "learning_rate": 0.00010479999999999999,
-      "loss": 3.6498,
+      "grad_norm": 2.915142297744751,
+      "learning_rate": 7.86e-05,
+      "loss": 3.6873,
       "step": 267
     },
     {
       "epoch": 0.4288,
-      "grad_norm": 1.0979737043380737,
-      "learning_rate": 0.0001052,
-      "loss": 3.5839,
+      "grad_norm": 2.7581217288970947,
+      "learning_rate": 7.89e-05,
+      "loss": 3.621,
       "step": 268
     },
     {
       "epoch": 0.4304,
-      "grad_norm": 2.4385807514190674,
-      "learning_rate": 0.00010559999999999998,
-      "loss": 3.6874,
+      "grad_norm": 0.830217719078064,
+      "learning_rate": 7.92e-05,
+      "loss": 3.7101,
       "step": 269
     },
     {
       "epoch": 0.432,
-      "grad_norm": 2.9847817420959473,
-      "learning_rate": 0.00010599999999999999,
-      "loss": 3.5842,
+      "grad_norm": 2.0595314502716064,
+      "learning_rate": 7.95e-05,
+      "loss": 3.6151,
       "step": 270
     },
     {
       "epoch": 0.4336,
-      "grad_norm": 6.495297431945801,
-      "learning_rate": 0.0001064,
-      "loss": 3.5658,
+      "grad_norm": 5.0488739013671875,
+      "learning_rate": 7.98e-05,
+      "loss": 3.6021,
       "step": 271
     },
     {
       "epoch": 0.4352,
-      "grad_norm": 3.128310441970825,
-      "learning_rate": 0.00010679999999999998,
-      "loss": 3.6348,
+      "grad_norm": 5.398313522338867,
+      "learning_rate": 8.01e-05,
+      "loss": 3.6869,
       "step": 272
     },
     {
       "epoch": 0.4368,
-      "grad_norm": 0.8451325297355652,
-      "learning_rate": 0.00010719999999999999,
-      "loss": 3.5906,
+      "grad_norm": 3.0345497131347656,
+      "learning_rate": 8.04e-05,
+      "loss": 3.6321,
       "step": 273
     },
     {
       "epoch": 0.4384,
-      "grad_norm": 1.0812450647354126,
-      "learning_rate": 0.0001076,
-      "loss": 3.5742,
+      "grad_norm": 5.039176940917969,
+      "learning_rate": 8.07e-05,
+      "loss": 3.6529,
       "step": 274
     },
     {
       "epoch": 0.44,
-      "grad_norm": 1.4351269006729126,
-      "learning_rate": 0.00010799999999999998,
-      "loss": 3.6606,
+      "grad_norm": 6.308064937591553,
+      "learning_rate": 8.1e-05,
+      "loss": 3.7383,
       "step": 275
     },
     {
       "epoch": 0.4416,
-      "grad_norm": 2.155651807785034,
-      "learning_rate": 0.00010839999999999999,
-      "loss": 3.6149,
+      "grad_norm": 2.592426300048828,
+      "learning_rate": 8.13e-05,
+      "loss": 3.6547,
       "step": 276
     },
     {
       "epoch": 0.4432,
-      "grad_norm": 1.8064038753509521,
-      "learning_rate": 0.0001088,
-      "loss": 3.5758,
+      "grad_norm": 1.9177695512771606,
+      "learning_rate": 8.16e-05,
+      "loss": 3.6161,
       "step": 277
     },
     {
       "epoch": 0.4448,
-      "grad_norm": 1.7468692064285278,
-      "learning_rate": 0.00010919999999999998,
-      "loss": 3.6068,
+      "grad_norm": 0.7630671858787537,
+      "learning_rate": 8.19e-05,
+      "loss": 3.6321,
       "step": 278
     },
     {
       "epoch": 0.4464,
-      "grad_norm": 1.2160580158233643,
-      "learning_rate": 0.0001096,
-      "loss": 3.5516,
+      "grad_norm": 1.042344093322754,
+      "learning_rate": 8.22e-05,
+      "loss": 3.5981,
       "step": 279
     },
     {
       "epoch": 0.448,
-      "grad_norm": 0.7980257868766785,
-      "learning_rate": 0.00010999999999999998,
-      "loss": 3.5646,
+      "grad_norm": 1.6470465660095215,
+      "learning_rate": 8.25e-05,
+      "loss": 3.6142,
       "step": 280
     },
     {
       "epoch": 0.4496,
-      "grad_norm": 4.852247714996338,
-      "learning_rate": 0.00011039999999999999,
-      "loss": 3.643,
+      "grad_norm": 1.3943268060684204,
+      "learning_rate": 8.28e-05,
+      "loss": 3.6655,
       "step": 281
     },
     {
       "epoch": 0.4512,
-      "grad_norm": 4.469863414764404,
-      "learning_rate": 0.0001108,
-      "loss": 3.5545,
+      "grad_norm": 1.1460942029953003,
+      "learning_rate": 8.31e-05,
+      "loss": 3.5895,
       "step": 282
     },
     {
       "epoch": 0.4528,
-      "grad_norm": 1.6534397602081299,
-      "learning_rate": 0.00011119999999999998,
-      "loss": 3.6189,
+      "grad_norm": 2.365217685699463,
+      "learning_rate": 8.34e-05,
+      "loss": 3.6698,
       "step": 283
     },
     {
       "epoch": 0.4544,
-      "grad_norm": 1.1284641027450562,
-      "learning_rate": 0.00011159999999999999,
-      "loss": 3.5333,
+      "grad_norm": 1.7076328992843628,
+      "learning_rate": 8.37e-05,
+      "loss": 3.586,
       "step": 284
     },
     {
       "epoch": 0.456,
-      "grad_norm": 2.752796173095703,
-      "learning_rate": 0.000112,
-      "loss": 3.69,
+      "grad_norm": 2.5929322242736816,
+      "learning_rate": 8.4e-05,
+      "loss": 3.7324,
       "step": 285
     },
     {
       "epoch": 0.4576,
-      "grad_norm": 0.7017075419425964,
-      "learning_rate": 0.00011239999999999998,
-      "loss": 3.6554,
+      "grad_norm": 2.4930028915405273,
+      "learning_rate": 8.43e-05,
+      "loss": 3.7213,
       "step": 286
     },
     {
       "epoch": 0.4592,
-      "grad_norm": 3.3214011192321777,
-      "learning_rate": 0.00011279999999999999,
-      "loss": 3.6023,
+      "grad_norm": 0.6821871399879456,
+      "learning_rate": 8.459999999999998e-05,
+      "loss": 3.624,
       "step": 287
     },
     {
       "epoch": 0.4608,
-      "grad_norm": 4.091536045074463,
-      "learning_rate": 0.0001132,
-      "loss": 3.5451,
+      "grad_norm": 0.7649940848350525,
+      "learning_rate": 8.489999999999999e-05,
+      "loss": 3.5646,
       "step": 288
     },
     {
       "epoch": 0.4624,
-      "grad_norm": 4.1135358810424805,
-      "learning_rate": 0.00011359999999999998,
-      "loss": 3.6137,
+      "grad_norm": 0.6352983713150024,
+      "learning_rate": 8.519999999999998e-05,
+      "loss": 3.6214,
       "step": 289
     },
     {
       "epoch": 0.464,
-      "grad_norm": 4.408985137939453,
-      "learning_rate": 0.00011399999999999999,
-      "loss": 3.5501,
+      "grad_norm": 1.2602449655532837,
+      "learning_rate": 8.549999999999999e-05,
+      "loss": 3.5684,
       "step": 290
     },
     {
       "epoch": 0.4656,
-      "grad_norm": 1.6145490407943726,
-      "learning_rate": 0.0001144,
-      "loss": 3.662,
+      "grad_norm": 1.568101167678833,
+      "learning_rate": 8.579999999999998e-05,
+      "loss": 3.7024,
       "step": 291
     },
     {
       "epoch": 0.4672,
-      "grad_norm": 1.9789752960205078,
-      "learning_rate": 0.00011479999999999999,
-      "loss": 3.493,
+      "grad_norm": 0.8837398290634155,
+      "learning_rate": 8.609999999999999e-05,
+      "loss": 3.5396,
       "step": 292
     },
     {
       "epoch": 0.4688,
-      "grad_norm": 2.3461720943450928,
-      "learning_rate": 0.0001152,
-      "loss": 3.5499,
+      "grad_norm": 2.6796021461486816,
+      "learning_rate": 8.639999999999999e-05,
+      "loss": 3.6308,
       "step": 293
     },
     {
       "epoch": 0.4704,
-      "grad_norm": 3.17932391166687,
-      "learning_rate": 0.0001156,
-      "loss": 3.5999,
+      "grad_norm": 2.063849925994873,
+      "learning_rate": 8.669999999999998e-05,
+      "loss": 3.6546,
       "step": 294
     },
     {
       "epoch": 0.472,
-      "grad_norm": 3.1920735836029053,
-      "learning_rate": 0.00011599999999999999,
-      "loss": 3.534,
+      "grad_norm": 0.7611846327781677,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 3.5734,
       "step": 295
     },
     {
       "epoch": 0.4736,
-      "grad_norm": 4.00572395324707,
-      "learning_rate": 0.0001164,
-      "loss": 3.5961,
+      "grad_norm": 0.967029333114624,
+      "learning_rate": 8.729999999999998e-05,
+      "loss": 3.6327,
       "step": 296
     },
     {
       "epoch": 0.4752,
-      "grad_norm": 2.458010673522949,
-      "learning_rate": 0.00011679999999999998,
-      "loss": 3.6061,
+      "grad_norm": 1.4702003002166748,
+      "learning_rate": 8.759999999999999e-05,
+      "loss": 3.667,
       "step": 297
     },
     {
       "epoch": 0.4768,
-      "grad_norm": 2.6859042644500732,
-      "learning_rate": 0.00011719999999999999,
-      "loss": 3.7096,
+      "grad_norm": 1.1918103694915771,
+      "learning_rate": 8.789999999999998e-05,
+      "loss": 3.7682,
       "step": 298
     },
     {
       "epoch": 0.4784,
-      "grad_norm": 1.6110202074050903,
-      "learning_rate": 0.0001176,
-      "loss": 3.7597,
+      "grad_norm": 1.5398286581039429,
+      "learning_rate": 8.819999999999999e-05,
+      "loss": 3.8027,
       "step": 299
     },
     {
       "epoch": 0.48,
-      "grad_norm": 2.9628493785858154,
-      "learning_rate": 0.00011799999999999998,
-      "loss": 3.6763,
-      "step": 300
-    },
-    {
-      "epoch": 0.48,
-      "eval_cer": 0.9768884282696217,
-      "eval_loss": 3.7717783451080322,
-      "eval_runtime": 158.6864,
-      "eval_samples_per_second": 19.762,
-      "eval_steps_per_second": 1.235,
-      "eval_wer": 1.0,
+      "grad_norm": 1.8616193532943726,
+      "learning_rate": 8.849999999999998e-05,
+      "loss": 3.7379,
       "step": 300
     },
     {
       "epoch": 0.4816,
-      "grad_norm": 30.372163772583008,
-      "learning_rate": 0.00011839999999999999,
-      "loss": 4.7027,
+      "grad_norm": 46.695960998535156,
+      "learning_rate": 8.879999999999999e-05,
+      "loss": 5.3609,
       "step": 301
     },
     {
       "epoch": 0.4832,
-      "grad_norm": 48.826873779296875,
-      "learning_rate": 0.0001188,
-      "loss": 4.0337,
+      "grad_norm": 27.103071212768555,
+      "learning_rate": 8.909999999999998e-05,
+      "loss": 4.2902,
       "step": 302
     },
     {
       "epoch": 0.4848,
-      "grad_norm": 1.4793564081192017,
-      "learning_rate": 0.00011919999999999998,
-      "loss": 3.6913,
+      "grad_norm": 3.6255853176116943,
+      "learning_rate": 8.939999999999999e-05,
+      "loss": 3.6986,
       "step": 303
     },
     {
       "epoch": 0.4864,
-      "grad_norm": 3.596323251724243,
-      "learning_rate": 0.0001196,
-      "loss": 3.8567,
+      "grad_norm": 7.375110149383545,
+      "learning_rate": 8.969999999999998e-05,
+      "loss": 3.8779,
       "step": 304
     },
     {
       "epoch": 0.488,
-      "grad_norm": 2.8520336151123047,
-      "learning_rate": 0.00011999999999999999,
-      "loss": 3.6953,
+      "grad_norm": 1.4208751916885376,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 3.7136,
       "step": 305
     },
     {
       "epoch": 0.4896,
-      "grad_norm": 5.386791706085205,
-      "learning_rate": 0.00012039999999999999,
-      "loss": 3.6839,
+      "grad_norm": 3.40026593208313,
+      "learning_rate": 9.029999999999999e-05,
+      "loss": 3.6894,
       "step": 306
     },
     {
       "epoch": 0.4912,
-      "grad_norm": 2.8725674152374268,
-      "learning_rate": 0.0001208,
-      "loss": 3.7514,
+      "grad_norm": 2.0953195095062256,
+      "learning_rate": 9.059999999999999e-05,
+      "loss": 3.7951,
       "step": 307
     },
     {
       "epoch": 0.4928,
-      "grad_norm": 6.234537124633789,
-      "learning_rate": 0.00012119999999999999,
-      "loss": 3.6282,
+      "grad_norm": 6.6124043464660645,
+      "learning_rate": 9.089999999999999e-05,
+      "loss": 3.6771,
       "step": 308
     },
     {
       "epoch": 0.4944,
-      "grad_norm": 3.3072447776794434,
-      "learning_rate": 0.00012159999999999999,
-      "loss": 3.6293,
+      "grad_norm": 5.1553425788879395,
+      "learning_rate": 9.12e-05,
+      "loss": 3.6848,
       "step": 309
     },
     {
       "epoch": 0.496,
-      "grad_norm": 1.3689359426498413,
-      "learning_rate": 0.000122,
-      "loss": 3.5943,
+      "grad_norm": 3.769348382949829,
+      "learning_rate": 9.149999999999999e-05,
+      "loss": 3.6674,
       "step": 310
     },
     {
       "epoch": 0.4976,
-      "grad_norm": 1.3273299932479858,
-      "learning_rate": 0.0001224,
-      "loss": 3.5541,
+      "grad_norm": 2.442067861557007,
+      "learning_rate": 9.18e-05,
+      "loss": 3.6276,
       "step": 311
     },
     {
       "epoch": 0.4992,
-      "grad_norm": 3.4486923217773438,
-      "learning_rate": 0.00012279999999999998,
-      "loss": 3.5964,
+      "grad_norm": 0.9911152124404907,
+      "learning_rate": 9.209999999999999e-05,
+      "loss": 3.6433,
       "step": 312
     },
     {
       "epoch": 0.5008,
-      "grad_norm": 2.501321792602539,
-      "learning_rate": 0.00012319999999999999,
-      "loss": 3.5374,
+      "grad_norm": 0.9690512418746948,
+      "learning_rate": 9.24e-05,
+      "loss": 3.5752,
       "step": 313
     },
     {
       "epoch": 0.5024,
-      "grad_norm": 5.377409934997559,
-      "learning_rate": 0.0001236,
-      "loss": 3.5793,
+      "grad_norm": 3.90602707862854,
+      "learning_rate": 9.269999999999999e-05,
+      "loss": 3.6274,
       "step": 314
     },
     {
       "epoch": 0.504,
-      "grad_norm": 4.240380764007568,
-      "learning_rate": 0.00012399999999999998,
-      "loss": 3.6035,
+      "grad_norm": 4.008577823638916,
+      "learning_rate": 9.3e-05,
+      "loss": 3.6627,
       "step": 315
     },
     {
       "epoch": 0.5056,
-      "grad_norm": 4.2587714195251465,
-      "learning_rate": 0.0001244,
-      "loss": 3.5728,
+      "grad_norm": 3.5601375102996826,
+      "learning_rate": 9.329999999999999e-05,
+      "loss": 3.6389,
       "step": 316
     },
     {
       "epoch": 0.5072,
-      "grad_norm": 1.2207915782928467,
-      "learning_rate": 0.00012479999999999997,
-      "loss": 3.5864,
+      "grad_norm": 0.9909217357635498,
+      "learning_rate": 9.36e-05,
+      "loss": 3.6401,
       "step": 317
     },
     {
       "epoch": 0.5088,
-      "grad_norm": 1.0006273984909058,
-      "learning_rate": 0.00012519999999999998,
-      "loss": 3.5532,
+      "grad_norm": 1.6349003314971924,
+      "learning_rate": 9.389999999999999e-05,
+      "loss": 3.6151,
       "step": 318
     },
     {
       "epoch": 0.5104,
-      "grad_norm": 2.789165496826172,
-      "learning_rate": 0.0001256,
-      "loss": 3.5095,
+      "grad_norm": 1.3834391832351685,
+      "learning_rate": 9.419999999999999e-05,
+      "loss": 3.5776,
       "step": 319
     },
     {
       "epoch": 0.512,
-      "grad_norm": 1.3580331802368164,
-      "learning_rate": 0.00012599999999999997,
-      "loss": 3.6278,
+      "grad_norm": 0.7489935755729675,
+      "learning_rate": 9.449999999999999e-05,
+      "loss": 3.6691,
       "step": 320
     },
     {
       "epoch": 0.5136,
-      "grad_norm": 2.3157308101654053,
-      "learning_rate": 0.00012639999999999998,
-      "loss": 3.5437,
+      "grad_norm": 0.6968175172805786,
+      "learning_rate": 9.479999999999999e-05,
+      "loss": 3.587,
       "step": 321
     },
     {
       "epoch": 0.5152,
-      "grad_norm": 1.8099067211151123,
-      "learning_rate": 0.0001268,
-      "loss": 3.519,
+      "grad_norm": 1.2773737907409668,
+      "learning_rate": 9.51e-05,
+      "loss": 3.5799,
       "step": 322
     },
     {
       "epoch": 0.5168,
-      "grad_norm": 2.988795042037964,
-      "learning_rate": 0.00012719999999999997,
-      "loss": 3.574,
+      "grad_norm": 3.2610504627227783,
+      "learning_rate": 9.539999999999999e-05,
+      "loss": 3.6406,
       "step": 323
     },
     {
       "epoch": 0.5184,
-      "grad_norm": 1.933131456375122,
-      "learning_rate": 0.00012759999999999998,
-      "loss": 3.4918,
+      "grad_norm": 1.130247712135315,
+      "learning_rate": 9.57e-05,
+      "loss": 3.5499,
       "step": 324
     },
     {
       "epoch": 0.52,
-      "grad_norm": 1.9587433338165283,
-      "learning_rate": 0.000128,
-      "loss": 3.5376,
+      "grad_norm": 0.9368891716003418,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 3.6001,
       "step": 325
     },
     {
       "epoch": 0.5216,
-      "grad_norm": 3.22497820854187,
-      "learning_rate": 0.00012839999999999998,
-      "loss": 3.5487,
+      "grad_norm": 1.5623962879180908,
+      "learning_rate": 9.63e-05,
+      "loss": 3.6115,
       "step": 326
     },
     {
       "epoch": 0.5232,
-      "grad_norm": 1.520359992980957,
-      "learning_rate": 0.0001288,
-      "loss": 3.4637,
+      "grad_norm": 1.9312142133712769,
+      "learning_rate": 9.659999999999999e-05,
+      "loss": 3.5575,
       "step": 327
     },
     {
       "epoch": 0.5248,
-      "grad_norm": 1.7252830266952515,
-      "learning_rate": 0.00012919999999999997,
-      "loss": 3.4904,
+      "grad_norm": 2.629648208618164,
+      "learning_rate": 9.69e-05,
+      "loss": 3.5637,
       "step": 328
     },
     {
       "epoch": 0.5264,
-      "grad_norm": 2.588486909866333,
-      "learning_rate": 0.00012959999999999998,
-      "loss": 3.4088,
+      "grad_norm": 2.997903823852539,
+      "learning_rate": 9.719999999999999e-05,
+      "loss": 3.4773,
       "step": 329
     },
     {
       "epoch": 0.528,
-      "grad_norm": 2.2423901557922363,
-      "learning_rate": 0.00013,
-      "loss": 3.4682,
+      "grad_norm": 2.4038596153259277,
+      "learning_rate": 9.75e-05,
+      "loss": 3.5326,
       "step": 330
     },
     {
       "epoch": 0.5296,
-      "grad_norm": 2.314682960510254,
-      "learning_rate": 0.00013039999999999997,
-      "loss": 3.5424,
+      "grad_norm": 1.7766273021697998,
+      "learning_rate": 9.779999999999999e-05,
+      "loss": 3.5888,
       "step": 331
     },
     {
       "epoch": 0.5312,
-      "grad_norm": 3.1224074363708496,
-      "learning_rate": 0.00013079999999999998,
-      "loss": 3.5728,
+      "grad_norm": 2.1525938510894775,
+      "learning_rate": 9.81e-05,
+      "loss": 3.6005,
       "step": 332
     },
     {
       "epoch": 0.5328,
-      "grad_norm": 1.9981073141098022,
-      "learning_rate": 0.0001312,
-      "loss": 3.4646,
+      "grad_norm": 1.953926920890808,
+      "learning_rate": 9.839999999999999e-05,
+      "loss": 3.5273,
       "step": 333
     },
     {
       "epoch": 0.5344,
-      "grad_norm": 0.9668769836425781,
-      "learning_rate": 0.00013159999999999997,
-      "loss": 3.5032,
+      "grad_norm": 1.1791383028030396,
+      "learning_rate": 9.87e-05,
+      "loss": 3.5676,
       "step": 334
     },
     {
       "epoch": 0.536,
-      "grad_norm": 2.059154987335205,
-      "learning_rate": 0.00013199999999999998,
-      "loss": 3.5411,
+      "grad_norm": 1.4245538711547852,
+      "learning_rate": 9.9e-05,
+      "loss": 3.606,
       "step": 335
     },
     {
       "epoch": 0.5376,
-      "grad_norm": 1.0898948907852173,
-      "learning_rate": 0.0001324,
-      "loss": 3.4957,
+      "grad_norm": 0.9623693823814392,
+      "learning_rate": 9.93e-05,
+      "loss": 3.5484,
       "step": 336
     },
     {
       "epoch": 0.5392,
-      "grad_norm": 2.375235080718994,
-      "learning_rate": 0.00013279999999999998,
-      "loss": 3.5179,
+      "grad_norm": 2.230302333831787,
+      "learning_rate": 9.96e-05,
+      "loss": 3.5715,
       "step": 337
     },
     {
       "epoch": 0.5408,
-      "grad_norm": 2.4042155742645264,
-      "learning_rate": 0.00013319999999999999,
-      "loss": 3.5158,
+      "grad_norm": 2.5997812747955322,
+      "learning_rate": 9.99e-05,
+      "loss": 3.5682,
       "step": 338
     },
     {
       "epoch": 0.5424,
-      "grad_norm": 2.0467703342437744,
-      "learning_rate": 0.0001336,
-      "loss": 3.5472,
+      "grad_norm": 0.998353123664856,
+      "learning_rate": 0.0001002,
+      "loss": 3.6058,
       "step": 339
     },
     {
       "epoch": 0.544,
-      "grad_norm": 1.2783809900283813,
-      "learning_rate": 0.00013399999999999998,
-      "loss": 3.4051,
+      "grad_norm": 0.9916491508483887,
+      "learning_rate": 0.0001005,
+      "loss": 3.4561,
       "step": 340
     },
     {
       "epoch": 0.5456,
-      "grad_norm": 1.2459137439727783,
-      "learning_rate": 0.0001344,
-      "loss": 3.383,
+      "grad_norm": 1.0415515899658203,
+      "learning_rate": 0.0001008,
+      "loss": 3.4528,
       "step": 341
     },
     {
       "epoch": 0.5472,
-      "grad_norm": 11.236316680908203,
-      "learning_rate": 0.00013479999999999997,
-      "loss": 3.8726,
+      "grad_norm": 10.178709983825684,
+      "learning_rate": 0.0001011,
+      "loss": 3.8938,
       "step": 342
     },
     {
       "epoch": 0.5488,
-      "grad_norm": 5.548703670501709,
-      "learning_rate": 0.00013519999999999998,
-      "loss": 3.5908,
+      "grad_norm": 8.032876014709473,
+      "learning_rate": 0.0001014,
+      "loss": 3.7116,
       "step": 343
     },
     {
       "epoch": 0.5504,
-      "grad_norm": 2.2428252696990967,
-      "learning_rate": 0.0001356,
-      "loss": 3.494,
+      "grad_norm": 4.835402488708496,
+      "learning_rate": 0.00010169999999999999,
+      "loss": 3.5765,
       "step": 344
     },
     {
       "epoch": 0.552,
-      "grad_norm": 1.2952911853790283,
-      "learning_rate": 0.00013599999999999997,
-      "loss": 3.5514,
+      "grad_norm": 4.175968170166016,
+      "learning_rate": 0.000102,
+      "loss": 3.6138,
       "step": 345
     },
     {
       "epoch": 0.5536,
-      "grad_norm": 2.755897283554077,
-      "learning_rate": 0.00013639999999999998,
-      "loss": 3.4938,
+      "grad_norm": 0.8417240977287292,
+      "learning_rate": 0.00010229999999999999,
+      "loss": 3.539,
       "step": 346
     },
     {
       "epoch": 0.5552,
-      "grad_norm": 4.549272537231445,
-      "learning_rate": 0.0001368,
-      "loss": 3.5529,
+      "grad_norm": 2.533202648162842,
+      "learning_rate": 0.0001026,
+      "loss": 3.5745,
       "step": 347
     },
     {
       "epoch": 0.5568,
-      "grad_norm": 5.1981048583984375,
-      "learning_rate": 0.00013719999999999997,
-      "loss": 3.4982,
+      "grad_norm": 3.826181650161743,
+      "learning_rate": 0.0001029,
+      "loss": 3.4985,
       "step": 348
     },
     {
       "epoch": 0.5584,
-      "grad_norm": 1.9575941562652588,
-      "learning_rate": 0.00013759999999999998,
-      "loss": 3.5279,
+      "grad_norm": 1.0143622159957886,
+      "learning_rate": 0.00010319999999999999,
+      "loss": 3.559,
       "step": 349
     },
     {
       "epoch": 0.56,
-      "grad_norm": 3.6745364665985107,
-      "learning_rate": 0.000138,
-      "loss": 3.628,
+      "grad_norm": 3.292059898376465,
+      "learning_rate": 0.00010349999999999998,
+      "loss": 3.6784,
       "step": 350
     },
     {
       "epoch": 0.5616,
-      "grad_norm": 15.095296859741211,
-      "learning_rate": 0.00013839999999999998,
-      "loss": 3.8532,
+      "grad_norm": 14.462305068969727,
+      "learning_rate": 0.00010379999999999999,
+      "loss": 3.8307,
       "step": 351
     },
     {
       "epoch": 0.5632,
-      "grad_norm": 16.068857192993164,
-      "learning_rate": 0.00013879999999999999,
-      "loss": 3.9237,
+      "grad_norm": 15.524165153503418,
+      "learning_rate": 0.00010409999999999998,
+      "loss": 3.8937,
       "step": 352
     },
     {
       "epoch": 0.5648,
-      "grad_norm": 16.43937873840332,
-      "learning_rate": 0.0001392,
-      "loss": 3.7986,
+      "grad_norm": 8.758255004882812,
+      "learning_rate": 0.00010439999999999999,
+      "loss": 3.7877,
       "step": 353
     },
     {
       "epoch": 0.5664,
-      "grad_norm": 16.25771141052246,
-      "learning_rate": 0.00013959999999999998,
-      "loss": 4.018,
+      "grad_norm": 12.151189804077148,
+      "learning_rate": 0.00010469999999999998,
+      "loss": 3.8868,
       "step": 354
     },
     {
       "epoch": 0.568,
-      "grad_norm": 20.026491165161133,
-      "learning_rate": 0.00014,
-      "loss": 4.2146,
+      "grad_norm": 15.403145790100098,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 4.0839,
       "step": 355
     },
     {
       "epoch": 0.5696,
-      "grad_norm": 5.759159564971924,
-      "learning_rate": 0.0001404,
-      "loss": 3.5998,
+      "grad_norm": 5.038736820220947,
+      "learning_rate": 0.00010529999999999998,
+      "loss": 3.5888,
       "step": 356
     },
     {
       "epoch": 0.5712,
-      "grad_norm": 7.332910537719727,
-      "learning_rate": 0.00014079999999999998,
-      "loss": 3.7159,
+      "grad_norm": 2.910677909851074,
+      "learning_rate": 0.00010559999999999998,
+      "loss": 3.7264,
       "step": 357
     },
     {
       "epoch": 0.5728,
-      "grad_norm": 1.8525800704956055,
-      "learning_rate": 0.0001412,
-      "loss": 3.6049,
+      "grad_norm": 2.463589668273926,
+      "learning_rate": 0.00010589999999999999,
+      "loss": 3.6665,
       "step": 358
     },
     {
       "epoch": 0.5744,
-      "grad_norm": 2.778653144836426,
-      "learning_rate": 0.00014159999999999997,
-      "loss": 3.5283,
+      "grad_norm": 4.513500690460205,
+      "learning_rate": 0.00010619999999999998,
+      "loss": 3.6046,
       "step": 359
     },
     {
       "epoch": 0.576,
-      "grad_norm": 1.6119922399520874,
-      "learning_rate": 0.00014199999999999998,
-      "loss": 3.5938,
+      "grad_norm": 2.1371867656707764,
+      "learning_rate": 0.00010649999999999999,
+      "loss": 3.6617,
       "step": 360
     },
     {
       "epoch": 0.5776,
-      "grad_norm": 2.705361843109131,
-      "learning_rate": 0.0001424,
-      "loss": 3.568,
+      "grad_norm": 2.018497943878174,
+      "learning_rate": 0.00010679999999999998,
+      "loss": 3.6356,
       "step": 361
     },
     {
       "epoch": 0.5792,
-      "grad_norm": 4.27663516998291,
-      "learning_rate": 0.00014279999999999997,
-      "loss": 3.5577,
+      "grad_norm": 2.9791243076324463,
+      "learning_rate": 0.00010709999999999999,
+      "loss": 3.5828,
       "step": 362
     },
     {
       "epoch": 0.5808,
-      "grad_norm": 3.09662127494812,
-      "learning_rate": 0.00014319999999999998,
-      "loss": 3.5939,
+      "grad_norm": 0.9730789065361023,
+      "learning_rate": 0.00010739999999999998,
+      "loss": 3.6317,
       "step": 363
     },
     {
       "epoch": 0.5824,
-      "grad_norm": 2.659717559814453,
-      "learning_rate": 0.0001436,
-      "loss": 3.5241,
+      "grad_norm": 0.9338732957839966,
+      "learning_rate": 0.00010769999999999999,
+      "loss": 3.5758,
       "step": 364
     },
     {
       "epoch": 0.584,
-      "grad_norm": 4.375011444091797,
-      "learning_rate": 0.00014399999999999998,
-      "loss": 3.4024,
+      "grad_norm": 2.041814088821411,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 3.4225,
       "step": 365
     },
     {
       "epoch": 0.5856,
-      "grad_norm": 2.0820505619049072,
-      "learning_rate": 0.00014439999999999999,
-      "loss": 3.4192,
+      "grad_norm": 0.7372158765792847,
+      "learning_rate": 0.00010829999999999999,
+      "loss": 3.4787,
       "step": 366
     },
     {
       "epoch": 0.5872,
-      "grad_norm": 3.44874906539917,
-      "learning_rate": 0.0001448,
-      "loss": 3.4785,
+      "grad_norm": 4.640085220336914,
+      "learning_rate": 0.00010859999999999998,
+      "loss": 3.566,
       "step": 367
     },
     {
       "epoch": 0.5888,
-      "grad_norm": 20.90206527709961,
-      "learning_rate": 0.00014519999999999998,
-      "loss": 3.7141,
+      "grad_norm": 10.43756103515625,
+      "learning_rate": 0.00010889999999999999,
+      "loss": 3.7959,
       "step": 368
     },
     {
       "epoch": 0.5904,
-      "grad_norm": 12.979022979736328,
-      "learning_rate": 0.0001456,
-      "loss": 3.9909,
+      "grad_norm": 12.211026191711426,
+      "learning_rate": 0.00010919999999999998,
+      "loss": 4.0038,
       "step": 369
     },
     {
       "epoch": 0.592,
-      "grad_norm": 4.532771110534668,
-      "learning_rate": 0.000146,
-      "loss": 3.5721,
+      "grad_norm": 2.702982187271118,
+      "learning_rate": 0.00010949999999999999,
+      "loss": 3.5885,
       "step": 370
     },
     {
       "epoch": 0.5936,
-      "grad_norm": 6.729887962341309,
-      "learning_rate": 0.00014639999999999998,
-      "loss": 3.5496,
+      "grad_norm": 3.37894868850708,
+      "learning_rate": 0.00010979999999999999,
+      "loss": 3.5248,
       "step": 371
     },
     {
       "epoch": 0.5952,
-      "grad_norm": 1.9874441623687744,
-      "learning_rate": 0.0001468,
-      "loss": 3.4444,
+      "grad_norm": 1.3367834091186523,
+      "learning_rate": 0.00011009999999999999,
+      "loss": 3.5093,
       "step": 372
     },
     {
       "epoch": 0.5968,
-      "grad_norm": 1.2600610256195068,
-      "learning_rate": 0.00014719999999999997,
-      "loss": 3.4629,
+      "grad_norm": 1.7772578001022339,
+      "learning_rate": 0.00011039999999999999,
+      "loss": 3.5247,
       "step": 373
     },
     {
       "epoch": 0.5984,
-      "grad_norm": 1.6816917657852173,
-      "learning_rate": 0.00014759999999999998,
-      "loss": 3.5387,
+      "grad_norm": 1.5332499742507935,
+      "learning_rate": 0.0001107,
+      "loss": 3.6032,
       "step": 374
     },
     {
       "epoch": 0.6,
-      "grad_norm": 3.2985401153564453,
-      "learning_rate": 0.000148,
-      "loss": 3.4601,
+      "grad_norm": 2.7413148880004883,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 3.5101,
       "step": 375
     },
     {
       "epoch": 0.6016,
-      "grad_norm": 2.27606463432312,
-      "learning_rate": 0.00014839999999999998,
-      "loss": 3.4595,
+      "grad_norm": 0.6245549917221069,
+      "learning_rate": 0.0001113,
+      "loss": 3.5173,
       "step": 376
     },
     {
       "epoch": 0.6032,
-      "grad_norm": 2.2081143856048584,
-      "learning_rate": 0.00014879999999999998,
-      "loss": 3.5109,
+      "grad_norm": 0.9614344239234924,
+      "learning_rate": 0.00011159999999999999,
+      "loss": 3.5536,
       "step": 377
     },
     {
       "epoch": 0.6048,
-      "grad_norm": 3.556088924407959,
-      "learning_rate": 0.0001492,
-      "loss": 3.3938,
+      "grad_norm": 1.8567408323287964,
+      "learning_rate": 0.0001119,
+      "loss": 3.4189,
       "step": 378
     },
     {
       "epoch": 0.6064,
-      "grad_norm": 1.6859062910079956,
-      "learning_rate": 0.00014959999999999998,
-      "loss": 3.4073,
+      "grad_norm": 1.1231341361999512,
+      "learning_rate": 0.00011219999999999999,
+      "loss": 3.4579,
       "step": 379
     },
     {
       "epoch": 0.608,
-      "grad_norm": 1.718527913093567,
-      "learning_rate": 0.00015,
-      "loss": 3.4326,
+      "grad_norm": 2.492093086242676,
+      "learning_rate": 0.0001125,
+      "loss": 3.4925,
       "step": 380
     },
     {
       "epoch": 0.6096,
-      "grad_norm": 7.6954522132873535,
-      "learning_rate": 0.00015039999999999997,
-      "loss": 3.4145,
+      "grad_norm": 8.354537963867188,
+      "learning_rate": 0.00011279999999999999,
+      "loss": 3.4855,
       "step": 381
     },
     {
       "epoch": 0.6112,
-      "grad_norm": 3.9803624153137207,
-      "learning_rate": 0.0001508,
-      "loss": 3.5334,
+      "grad_norm": 3.4800968170166016,
+      "learning_rate": 0.00011309999999999998,
+      "loss": 3.5702,
       "step": 382
     },
     {
       "epoch": 0.6128,
-      "grad_norm": 4.626428127288818,
-      "learning_rate": 0.0001512,
-      "loss": 3.5431,
+      "grad_norm": 2.3893015384674072,
+      "learning_rate": 0.00011339999999999999,
+      "loss": 3.5403,
       "step": 383
     },
     {
       "epoch": 0.6144,
-      "grad_norm": 1.796976923942566,
-      "learning_rate": 0.00015159999999999997,
-      "loss": 3.3876,
+      "grad_norm": 1.4556490182876587,
+      "learning_rate": 0.00011369999999999999,
+      "loss": 3.4286,
       "step": 384
     },
     {
       "epoch": 0.616,
-      "grad_norm": 1.869494080543518,
-      "learning_rate": 0.000152,
-      "loss": 3.4063,
+      "grad_norm": 0.9344227910041809,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 3.4456,
       "step": 385
     },
     {
       "epoch": 0.6176,
-      "grad_norm": 1.6276867389678955,
-      "learning_rate": 0.0001524,
-      "loss": 3.435,
+      "grad_norm": 1.8487316370010376,
+      "learning_rate": 0.00011429999999999999,
+      "loss": 3.5025,
       "step": 386
     },
     {
       "epoch": 0.6192,
-      "grad_norm": 1.9487453699111938,
-      "learning_rate": 0.00015279999999999997,
-      "loss": 3.4772,
+      "grad_norm": 2.2494537830352783,
+      "learning_rate": 0.0001146,
+      "loss": 3.5277,
       "step": 387
     },
     {
       "epoch": 0.6208,
-      "grad_norm": 3.8081955909729004,
-      "learning_rate": 0.0001532,
-      "loss": 3.4428,
+      "grad_norm": 1.3514158725738525,
+      "learning_rate": 0.00011489999999999999,
+      "loss": 3.4763,
       "step": 388
     },
     {
       "epoch": 0.6224,
-      "grad_norm": 1.7065116167068481,
-      "learning_rate": 0.0001536,
-      "loss": 3.4732,
+      "grad_norm": 1.518999695777893,
+      "learning_rate": 0.0001152,
+      "loss": 3.4946,
       "step": 389
     },
     {
       "epoch": 0.624,
-      "grad_norm": 4.1957292556762695,
-      "learning_rate": 0.00015399999999999998,
-      "loss": 3.4299,
+      "grad_norm": 1.8156572580337524,
+      "learning_rate": 0.00011549999999999999,
+      "loss": 3.4156,
       "step": 390
     },
     {
       "epoch": 0.6256,
-      "grad_norm": 2.8493101596832275,
-      "learning_rate": 0.0001544,
-      "loss": 3.3279,
+      "grad_norm": 1.6927382946014404,
+      "learning_rate": 0.0001158,
+      "loss": 3.3432,
       "step": 391
     },
     {
       "epoch": 0.6272,
-      "grad_norm": 1.7241597175598145,
-      "learning_rate": 0.0001548,
-      "loss": 3.3818,
+      "grad_norm": 1.9768507480621338,
+      "learning_rate": 0.00011609999999999999,
+      "loss": 3.4614,
       "step": 392
     },
     {
       "epoch": 0.6288,
-      "grad_norm": 1.3103301525115967,
-      "learning_rate": 0.00015519999999999998,
-      "loss": 3.4189,
+      "grad_norm": 2.5529818534851074,
+      "learning_rate": 0.0001164,
+      "loss": 3.4839,
       "step": 393
     },
     {
       "epoch": 0.6304,
-      "grad_norm": 1.0086480379104614,
-      "learning_rate": 0.00015560000000000001,
-      "loss": 3.3741,
+      "grad_norm": 1.4785748720169067,
+      "learning_rate": 0.00011669999999999999,
+      "loss": 3.4267,
       "step": 394
     },
     {
       "epoch": 0.632,
-      "grad_norm": 2.4616684913635254,
-      "learning_rate": 0.000156,
-      "loss": 3.3846,
+      "grad_norm": 0.9147313237190247,
+      "learning_rate": 0.000117,
+      "loss": 3.4296,
       "step": 395
     },
     {
       "epoch": 0.6336,
-      "grad_norm": 4.7007527351379395,
-      "learning_rate": 0.00015639999999999998,
-      "loss": 3.4063,
+      "grad_norm": 1.8749762773513794,
+      "learning_rate": 0.00011729999999999999,
+      "loss": 3.3974,
       "step": 396
     },
     {
       "epoch": 0.6352,
-      "grad_norm": 2.0122146606445312,
-      "learning_rate": 0.00015679999999999996,
-      "loss": 3.4581,
+      "grad_norm": 2.243661403656006,
+      "learning_rate": 0.0001176,
+      "loss": 3.4929,
       "step": 397
     },
     {
       "epoch": 0.6368,
-      "grad_norm": 1.6844937801361084,
-      "learning_rate": 0.0001572,
-      "loss": 3.4897,
+      "grad_norm": 3.580742120742798,
+      "learning_rate": 0.00011789999999999999,
+      "loss": 3.5545,
       "step": 398
     },
     {
       "epoch": 0.6384,
-      "grad_norm": 2.5921356678009033,
-      "learning_rate": 0.00015759999999999998,
-      "loss": 3.4199,
+      "grad_norm": 2.198838472366333,
+      "learning_rate": 0.0001182,
+      "loss": 3.4712,
       "step": 399
     },
     {
       "epoch": 0.64,
       "grad_norm": NaN,
-      "learning_rate": 0.00015759999999999998,
-      "loss": 3.1272,
-      "step": 400
-    },
-    {
-      "epoch": 0.64,
-      "eval_cer": 0.982628452983617,
-      "eval_loss": 3.464660882949829,
-      "eval_runtime": 157.7514,
-      "eval_samples_per_second": 19.879,
-      "eval_steps_per_second": 1.242,
-      "eval_wer": 1.0,
+      "learning_rate": 0.0001182,
+      "loss": 3.1738,
       "step": 400
     },
     {
       "epoch": 0.6416,
-      "grad_norm": 14.215404510498047,
-      "learning_rate": 0.00015799999999999996,
-      "loss": 3.8237,
+      "grad_norm": 16.711830139160156,
+      "learning_rate": 0.0001185,
+      "loss": 3.9305,
       "step": 401
     },
     {
       "epoch": 0.6432,
-      "grad_norm": 11.251092910766602,
-      "learning_rate": 0.0001584,
-      "loss": 3.6592,
+      "grad_norm": 15.2301607131958,
+      "learning_rate": 0.0001188,
+      "loss": 3.812,
       "step": 402
     },
     {
       "epoch": 0.6448,
-      "grad_norm": 23.282085418701172,
-      "learning_rate": 0.00015879999999999998,
-      "loss": 4.4136,
+      "grad_norm": 31.35503387451172,
+      "learning_rate": 0.0001191,
+      "loss": 4.7454,
       "step": 403
     },
     {
       "epoch": 0.6464,
-      "grad_norm": 4.128000259399414,
-      "learning_rate": 0.00015919999999999997,
-      "loss": 3.5762,
+      "grad_norm": 10.81604290008545,
+      "learning_rate": 0.0001194,
+      "loss": 3.7583,
       "step": 404
     },
     {
       "epoch": 0.648,
-      "grad_norm": 3.7951207160949707,
-      "learning_rate": 0.0001596,
-      "loss": 3.6869,
+      "grad_norm": 11.12818717956543,
+      "learning_rate": 0.0001197,
+      "loss": 3.8342,
       "step": 405
     },
     {
       "epoch": 0.6496,
-      "grad_norm": 2.6995790004730225,
-      "learning_rate": 0.00015999999999999999,
-      "loss": 3.6114,
+      "grad_norm": 5.6629767417907715,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 3.707,
       "step": 406
     },
     {
       "epoch": 0.6512,
-      "grad_norm": 8.775090217590332,
-      "learning_rate": 0.00016039999999999997,
-      "loss": 3.5307,
+      "grad_norm": 5.830606937408447,
+      "learning_rate": 0.0001203,
+      "loss": 3.4846,
       "step": 407
     },
     {
       "epoch": 0.6528,
-      "grad_norm": 7.154327392578125,
-      "learning_rate": 0.0001608,
-      "loss": 3.568,
+      "grad_norm": 4.803534030914307,
+      "learning_rate": 0.00012059999999999999,
+      "loss": 3.5607,
       "step": 408
     },
     {
       "epoch": 0.6544,
-      "grad_norm": 5.456339359283447,
-      "learning_rate": 0.0001612,
-      "loss": 3.5659,
+      "grad_norm": 4.736687660217285,
+      "learning_rate": 0.0001209,
+      "loss": 3.5705,
       "step": 409
     },
     {
       "epoch": 0.656,
-      "grad_norm": 5.459187030792236,
-      "learning_rate": 0.00016159999999999997,
-      "loss": 3.5469,
+      "grad_norm": 5.9123101234436035,
+      "learning_rate": 0.00012119999999999999,
+      "loss": 3.5491,
       "step": 410
     },
     {
       "epoch": 0.6576,
-      "grad_norm": 4.668329238891602,
-      "learning_rate": 0.000162,
-      "loss": 3.5329,
+      "grad_norm": 4.854342460632324,
+      "learning_rate": 0.0001215,
+      "loss": 3.5639,
       "step": 411
     },
     {
       "epoch": 0.6592,
-      "grad_norm": 2.4972665309906006,
-      "learning_rate": 0.0001624,
-      "loss": 3.4671,
+      "grad_norm": 2.654384136199951,
+      "learning_rate": 0.00012179999999999999,
+      "loss": 3.519,
       "step": 412
     },
     {
       "epoch": 0.6608,
-      "grad_norm": 2.261772394180298,
-      "learning_rate": 0.00016279999999999997,
-      "loss": 3.4179,
+      "grad_norm": 2.97538161277771,
+      "learning_rate": 0.00012209999999999999,
+      "loss": 3.4907,
       "step": 413
     },
     {
       "epoch": 0.6624,
-      "grad_norm": 1.6107017993927002,
-      "learning_rate": 0.0001632,
-      "loss": 3.411,
+      "grad_norm": 1.8297629356384277,
+      "learning_rate": 0.0001224,
+      "loss": 3.4962,
       "step": 414
     },
     {
       "epoch": 0.664,
-      "grad_norm": 3.2576379776000977,
-      "learning_rate": 0.0001636,
-      "loss": 3.4519,
+      "grad_norm": 3.1731834411621094,
+      "learning_rate": 0.00012269999999999997,
+      "loss": 3.5136,
       "step": 415
     },
     {
       "epoch": 0.6656,
-      "grad_norm": 6.433845520019531,
-      "learning_rate": 0.00016399999999999997,
-      "loss": 3.5618,
+      "grad_norm": 7.233218669891357,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 3.6006,
       "step": 416
     },
     {
       "epoch": 0.6672,
-      "grad_norm": 3.0946199893951416,
-      "learning_rate": 0.0001644,
-      "loss": 3.4222,
+      "grad_norm": 3.845116376876831,
+      "learning_rate": 0.0001233,
+      "loss": 3.4742,
       "step": 417
     },
     {
       "epoch": 0.6688,
-      "grad_norm": 4.258899688720703,
-      "learning_rate": 0.0001648,
-      "loss": 3.5111,
+      "grad_norm": 5.4477152824401855,
+      "learning_rate": 0.0001236,
+      "loss": 3.5832,
       "step": 418
     },
     {
       "epoch": 0.6704,
-      "grad_norm": 4.503801345825195,
-      "learning_rate": 0.00016519999999999998,
-      "loss": 3.4504,
+      "grad_norm": 5.551750659942627,
+      "learning_rate": 0.00012389999999999998,
+      "loss": 3.5168,
       "step": 419
     },
     {
       "epoch": 0.672,
-      "grad_norm": 0.8702706694602966,
-      "learning_rate": 0.0001656,
-      "loss": 3.3714,
+      "grad_norm": 1.164810299873352,
+      "learning_rate": 0.00012419999999999998,
+      "loss": 3.4397,
       "step": 420
     },
     {
       "epoch": 0.6736,
-      "grad_norm": 1.4191309213638306,
-      "learning_rate": 0.000166,
-      "loss": 3.3781,
+      "grad_norm": 1.6991772651672363,
+      "learning_rate": 0.0001245,
+      "loss": 3.4452,
       "step": 421
     },
     {
       "epoch": 0.6752,
-      "grad_norm": 1.344421148300171,
-      "learning_rate": 0.00016639999999999998,
-      "loss": 3.3594,
+      "grad_norm": 1.1151283979415894,
+      "learning_rate": 0.00012479999999999997,
+      "loss": 3.4537,
       "step": 422
     },
     {
       "epoch": 0.6768,
-      "grad_norm": 1.4346777200698853,
-      "learning_rate": 0.0001668,
-      "loss": 3.3227,
+      "grad_norm": 1.9064522981643677,
+      "learning_rate": 0.00012509999999999998,
+      "loss": 3.4217,
       "step": 423
     },
     {
       "epoch": 0.6784,
-      "grad_norm": 3.5542514324188232,
-      "learning_rate": 0.0001672,
-      "loss": 3.3607,
+      "grad_norm": 4.329729080200195,
+      "learning_rate": 0.00012539999999999999,
+      "loss": 3.4353,
       "step": 424
     },
     {
       "epoch": 0.68,
-      "grad_norm": 2.4777677059173584,
-      "learning_rate": 0.00016759999999999998,
-      "loss": 3.3752,
+      "grad_norm": 3.2471909523010254,
+      "learning_rate": 0.0001257,
+      "loss": 3.4603,
       "step": 425
     },
     {
       "epoch": 0.6816,
-      "grad_norm": 1.9810409545898438,
-      "learning_rate": 0.000168,
-      "loss": 3.3176,
+      "grad_norm": 2.613004207611084,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 3.4278,
       "step": 426
     },
     {
       "epoch": 0.6832,
-      "grad_norm": 1.3213425874710083,
-      "learning_rate": 0.0001684,
-      "loss": 3.3277,
+      "grad_norm": 0.7707146406173706,
+      "learning_rate": 0.00012629999999999998,
+      "loss": 3.4114,
       "step": 427
     },
     {
       "epoch": 0.6848,
-      "grad_norm": 1.7977039813995361,
-      "learning_rate": 0.00016879999999999998,
-      "loss": 3.2458,
+      "grad_norm": 1.4232515096664429,
+      "learning_rate": 0.0001266,
+      "loss": 3.3683,
       "step": 428
     },
     {
       "epoch": 0.6864,
-      "grad_norm": 1.8572102785110474,
-      "learning_rate": 0.00016919999999999997,
-      "loss": 3.2321,
+      "grad_norm": 2.6852142810821533,
+      "learning_rate": 0.0001269,
+      "loss": 3.3575,
       "step": 429
     },
     {
       "epoch": 0.688,
-      "grad_norm": 1.9749292135238647,
-      "learning_rate": 0.0001696,
-      "loss": 3.3227,
+      "grad_norm": 2.602998733520508,
+      "learning_rate": 0.00012719999999999997,
+      "loss": 3.4309,
       "step": 430
     },
     {
       "epoch": 0.6896,
-      "grad_norm": 1.989938735961914,
-      "learning_rate": 0.00016999999999999999,
-      "loss": 3.2361,
+      "grad_norm": 2.8761863708496094,
+      "learning_rate": 0.00012749999999999998,
+      "loss": 3.3435,
       "step": 431
     },
     {
       "epoch": 0.6912,
-      "grad_norm": 1.1900919675827026,
-      "learning_rate": 0.00017039999999999997,
-      "loss": 3.1931,
+      "grad_norm": 0.779827356338501,
+      "learning_rate": 0.0001278,
+      "loss": 3.3467,
       "step": 432
     },
     {
       "epoch": 0.6928,
-      "grad_norm": 1.588123083114624,
-      "learning_rate": 0.0001708,
-      "loss": 3.1711,
+      "grad_norm": 1.6401891708374023,
+      "learning_rate": 0.0001281,
+      "loss": 3.3293,
       "step": 433
     },
     {
       "epoch": 0.6944,
-      "grad_norm": 2.484400749206543,
-      "learning_rate": 0.0001712,
-      "loss": 3.3573,
+      "grad_norm": 1.1276079416275024,
+      "learning_rate": 0.00012839999999999998,
+      "loss": 3.4566,
       "step": 434
     },
     {
       "epoch": 0.696,
-      "grad_norm": 1.7051153182983398,
-      "learning_rate": 0.00017159999999999997,
-      "loss": 3.2167,
+      "grad_norm": 1.1479798555374146,
+      "learning_rate": 0.00012869999999999998,
+      "loss": 3.3803,
       "step": 435
     },
     {
       "epoch": 0.6976,
-      "grad_norm": 2.2069146633148193,
-      "learning_rate": 0.000172,
-      "loss": 3.393,
+      "grad_norm": 1.1874502897262573,
+      "learning_rate": 0.000129,
+      "loss": 3.4799,
       "step": 436
     },
     {
       "epoch": 0.6992,
-      "grad_norm": 1.0689940452575684,
-      "learning_rate": 0.0001724,
-      "loss": 3.3359,
+      "grad_norm": 1.1631622314453125,
+      "learning_rate": 0.0001293,
+      "loss": 3.4652,
       "step": 437
     },
     {
       "epoch": 0.7008,
-      "grad_norm": 2.242213487625122,
-      "learning_rate": 0.00017279999999999997,
-      "loss": 3.2326,
+      "grad_norm": 1.8987886905670166,
+      "learning_rate": 0.00012959999999999998,
+      "loss": 3.3634,
       "step": 438
     },
     {
       "epoch": 0.7024,
-      "grad_norm": 1.2001672983169556,
-      "learning_rate": 0.00017319999999999998,
-      "loss": 3.1571,
+      "grad_norm": 1.0175995826721191,
+      "learning_rate": 0.00012989999999999999,
+      "loss": 3.3244,
       "step": 439
     },
     {
       "epoch": 0.704,
-      "grad_norm": 1.493927001953125,
-      "learning_rate": 0.0001736,
-      "loss": 3.281,
+      "grad_norm": 2.3897902965545654,
+      "learning_rate": 0.0001302,
+      "loss": 3.4436,
       "step": 440
     },
     {
       "epoch": 0.7056,
-      "grad_norm": 1.6951879262924194,
-      "learning_rate": 0.00017399999999999997,
-      "loss": 3.2278,
+      "grad_norm": 2.532571792602539,
+      "learning_rate": 0.0001305,
+      "loss": 3.4136,
       "step": 441
     },
     {
       "epoch": 0.7072,
-      "grad_norm": 1.2256603240966797,
-      "learning_rate": 0.00017439999999999998,
-      "loss": 3.2015,
+      "grad_norm": 1.2601231336593628,
+      "learning_rate": 0.00013079999999999998,
+      "loss": 3.3457,
       "step": 442
     },
     {
       "epoch": 0.7088,
-      "grad_norm": 2.6825082302093506,
-      "learning_rate": 0.0001748,
-      "loss": 3.3138,
+      "grad_norm": 2.287309408187866,
+      "learning_rate": 0.0001311,
+      "loss": 3.4499,
       "step": 443
     },
     {
       "epoch": 0.7104,
-      "grad_norm": 2.056058645248413,
-      "learning_rate": 0.00017519999999999998,
-      "loss": 3.1577,
+      "grad_norm": 2.2662341594696045,
+      "learning_rate": 0.0001314,
+      "loss": 3.3269,
       "step": 444
     },
     {
       "epoch": 0.712,
-      "grad_norm": 2.27656888961792,
-      "learning_rate": 0.00017559999999999999,
-      "loss": 3.2605,
+      "grad_norm": 3.020737886428833,
+      "learning_rate": 0.00013169999999999998,
+      "loss": 3.4164,
       "step": 445
     },
     {
       "epoch": 0.7136,
-      "grad_norm": 1.663690209388733,
-      "learning_rate": 0.000176,
-      "loss": 3.1057,
+      "grad_norm": 2.3452391624450684,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 3.3253,
       "step": 446
     },
     {
       "epoch": 0.7152,
-      "grad_norm": 1.4355462789535522,
-      "learning_rate": 0.00017639999999999998,
-      "loss": 3.3227,
+      "grad_norm": 2.1812047958374023,
+      "learning_rate": 0.0001323,
+      "loss": 3.462,
       "step": 447
     },
     {
       "epoch": 0.7168,
-      "grad_norm": 1.0722546577453613,
-      "learning_rate": 0.0001768,
-      "loss": 3.1746,
+      "grad_norm": 1.294961929321289,
+      "learning_rate": 0.0001326,
+      "loss": 3.3657,
       "step": 448
     },
     {
       "epoch": 0.7184,
-      "grad_norm": 3.61966872215271,
-      "learning_rate": 0.0001772,
-      "loss": 3.2868,
+      "grad_norm": 4.165415287017822,
+      "learning_rate": 0.00013289999999999998,
+      "loss": 3.4706,
       "step": 449
     },
     {
       "epoch": 0.72,
-      "grad_norm": 3.237584114074707,
-      "learning_rate": 0.00017759999999999998,
-      "loss": 3.5749,
+      "grad_norm": 3.625424861907959,
+      "learning_rate": 0.00013319999999999999,
+      "loss": 3.6857,
       "step": 450
     },
     {
       "epoch": 0.7216,
-      "grad_norm": 42.27267074584961,
-      "learning_rate": 0.000178,
-      "loss": 5.2106,
+      "grad_norm": 42.7130012512207,
+      "learning_rate": 0.0001335,
+      "loss": 5.3364,
       "step": 451
     },
     {
       "epoch": 0.7232,
-      "grad_norm": 17.318639755249023,
-      "learning_rate": 0.0001784,
-      "loss": 3.8636,
+      "grad_norm": 18.547584533691406,
+      "learning_rate": 0.0001338,
+      "loss": 4.0571,
       "step": 452
     },
     {
       "epoch": 0.7248,
-      "grad_norm": 8.741928100585938,
-      "learning_rate": 0.00017879999999999998,
-      "loss": 3.4172,
+      "grad_norm": 11.535635948181152,
+      "learning_rate": 0.00013409999999999998,
+      "loss": 3.7247,
       "step": 453
     },
     {
       "epoch": 0.7264,
-      "grad_norm": 5.296689510345459,
-      "learning_rate": 0.0001792,
-      "loss": 3.4227,
+      "grad_norm": 6.823582649230957,
+      "learning_rate": 0.0001344,
+      "loss": 3.6886,
       "step": 454
     },
     {
       "epoch": 0.728,
-      "grad_norm": 1.3165295124053955,
-      "learning_rate": 0.0001796,
-      "loss": 3.2189,
+      "grad_norm": 2.5420644283294678,
+      "learning_rate": 0.0001347,
+      "loss": 3.5274,
       "step": 455
     },
     {
       "epoch": 0.7296,
-      "grad_norm": 1.799149990081787,
-      "learning_rate": 0.00017999999999999998,
-      "loss": 3.1344,
+      "grad_norm": 0.9652547240257263,
+      "learning_rate": 0.000135,
+      "loss": 3.4573,
       "step": 456
     },
     {
       "epoch": 0.7312,
-      "grad_norm": 1.491166591644287,
-      "learning_rate": 0.0001804,
-      "loss": 3.1914,
+      "grad_norm": 1.7442926168441772,
+      "learning_rate": 0.00013529999999999998,
+      "loss": 3.4887,
       "step": 457
     },
     {
       "epoch": 0.7328,
-      "grad_norm": 3.692824363708496,
-      "learning_rate": 0.00018079999999999998,
-      "loss": 3.1351,
+      "grad_norm": 5.671680927276611,
+      "learning_rate": 0.0001356,
+      "loss": 3.4587,
       "step": 458
     },
     {
       "epoch": 0.7344,
-      "grad_norm": 2.3572800159454346,
-      "learning_rate": 0.00018119999999999999,
-      "loss": 3.0732,
+      "grad_norm": 5.160175800323486,
+      "learning_rate": 0.0001359,
+      "loss": 3.4867,
       "step": 459
     },
     {
       "epoch": 0.736,
-      "grad_norm": 1.9129135608673096,
-      "learning_rate": 0.00018159999999999997,
-      "loss": 3.1997,
+      "grad_norm": 4.00103235244751,
+      "learning_rate": 0.0001362,
+      "loss": 3.5526,
       "step": 460
     },
     {
       "epoch": 0.7376,
-      "grad_norm": 1.8856866359710693,
-      "learning_rate": 0.00018199999999999998,
-      "loss": 3.0315,
+      "grad_norm": 4.9756975173950195,
+      "learning_rate": 0.00013649999999999998,
+      "loss": 3.4155,
       "step": 461
     },
     {
       "epoch": 0.7392,
-      "grad_norm": 2.009660243988037,
-      "learning_rate": 0.0001824,
-      "loss": 3.0865,
+      "grad_norm": 2.167109489440918,
+      "learning_rate": 0.0001368,
+      "loss": 3.4686,
       "step": 462
     },
     {
       "epoch": 0.7408,
-      "grad_norm": 2.5454518795013428,
-      "learning_rate": 0.00018279999999999997,
-      "loss": 2.9954,
+      "grad_norm": 1.0356065034866333,
+      "learning_rate": 0.0001371,
+      "loss": 3.4003,
       "step": 463
     },
     {
       "epoch": 0.7424,
-      "grad_norm": 0.8470408916473389,
-      "learning_rate": 0.00018319999999999998,
-      "loss": 2.9569,
+      "grad_norm": 1.7771834135055542,
+      "learning_rate": 0.0001374,
+      "loss": 3.3392,
       "step": 464
     },
     {
       "epoch": 0.744,
-      "grad_norm": 1.2759836912155151,
-      "learning_rate": 0.0001836,
-      "loss": 2.9149,
+      "grad_norm": 1.7832553386688232,
+      "learning_rate": 0.00013769999999999999,
+      "loss": 3.3132,
       "step": 465
     },
     {
       "epoch": 0.7456,
-      "grad_norm": 0.9253954291343689,
-      "learning_rate": 0.00018399999999999997,
-      "loss": 2.8104,
+      "grad_norm": 2.8095998764038086,
+      "learning_rate": 0.000138,
+      "loss": 3.2934,
       "step": 466
     },
     {
       "epoch": 0.7472,
-      "grad_norm": 1.043556809425354,
-      "learning_rate": 0.00018439999999999998,
-      "loss": 2.8224,
+      "grad_norm": 2.455730438232422,
+      "learning_rate": 0.0001383,
+      "loss": 3.3035,
       "step": 467
     },
     {
       "epoch": 0.7488,
-      "grad_norm": 2.264782190322876,
-      "learning_rate": 0.0001848,
-      "loss": 2.7933,
+      "grad_norm": 2.5177114009857178,
+      "learning_rate": 0.0001386,
+      "loss": 3.3005,
       "step": 468
     },
     {
       "epoch": 0.7504,
-      "grad_norm": 1.1183011531829834,
-      "learning_rate": 0.00018519999999999998,
-      "loss": 2.8228,
+      "grad_norm": 0.8732132315635681,
+      "learning_rate": 0.0001389,
+      "loss": 3.2456,
       "step": 469
     },
     {
       "epoch": 0.752,
-      "grad_norm": 0.8123898506164551,
-      "learning_rate": 0.00018559999999999998,
-      "loss": 2.7909,
+      "grad_norm": 1.003593921661377,
+      "learning_rate": 0.0001392,
+      "loss": 3.2672,
       "step": 470
     },
     {
       "epoch": 0.7536,
-      "grad_norm": 0.9366122484207153,
-      "learning_rate": 0.000186,
-      "loss": 2.7708,
+      "grad_norm": 1.1762638092041016,
+      "learning_rate": 0.0001395,
+      "loss": 3.2424,
       "step": 471
     },
     {
       "epoch": 0.7552,
-      "grad_norm": 1.1672213077545166,
-      "learning_rate": 0.00018639999999999998,
-      "loss": 2.8047,
+      "grad_norm": 1.4990063905715942,
+      "learning_rate": 0.00013979999999999998,
+      "loss": 3.2615,
       "step": 472
     },
     {
       "epoch": 0.7568,
-      "grad_norm": 2.68989634513855,
-      "learning_rate": 0.0001868,
-      "loss": 2.7845,
+      "grad_norm": 2.9979848861694336,
+      "learning_rate": 0.0001401,
+      "loss": 3.3328,
       "step": 473
     },
     {
       "epoch": 0.7584,
-      "grad_norm": 1.9162228107452393,
-      "learning_rate": 0.0001872,
-      "loss": 2.6011,
+      "grad_norm": 1.6674293279647827,
+      "learning_rate": 0.0001404,
+      "loss": 3.1763,
       "step": 474
     },
     {
       "epoch": 0.76,
-      "grad_norm": 2.0587878227233887,
-      "learning_rate": 0.00018759999999999998,
-      "loss": 2.7066,
+      "grad_norm": 1.2908185720443726,
+      "learning_rate": 0.00014069999999999998,
+      "loss": 3.2605,
       "step": 475
     },
     {
       "epoch": 0.7616,
-      "grad_norm": 0.9546549320220947,
-      "learning_rate": 0.000188,
-      "loss": 2.7525,
+      "grad_norm": 0.7701535820960999,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 3.2527,
       "step": 476
     },
     {
       "epoch": 0.7632,
-      "grad_norm": 1.305992841720581,
-      "learning_rate": 0.00018839999999999997,
-      "loss": 2.8766,
+      "grad_norm": 1.2732120752334595,
+      "learning_rate": 0.0001413,
+      "loss": 3.393,
       "step": 477
     },
     {
       "epoch": 0.7648,
-      "grad_norm": 1.25698721408844,
-      "learning_rate": 0.00018879999999999998,
-      "loss": 2.7792,
+      "grad_norm": 1.0643001794815063,
+      "learning_rate": 0.00014159999999999997,
+      "loss": 3.2402,
       "step": 478
     },
     {
       "epoch": 0.7664,
-      "grad_norm": 1.4074276685714722,
-      "learning_rate": 0.0001892,
-      "loss": 2.5459,
+      "grad_norm": 1.2950279712677002,
+      "learning_rate": 0.00014189999999999998,
+      "loss": 3.1784,
       "step": 479
     },
     {
       "epoch": 0.768,
-      "grad_norm": 1.1359269618988037,
-      "learning_rate": 0.00018959999999999997,
-      "loss": 2.5343,
+      "grad_norm": 1.4429744482040405,
+      "learning_rate": 0.0001422,
+      "loss": 3.1181,
       "step": 480
     },
     {
       "epoch": 0.7696,
-      "grad_norm": 1.2957308292388916,
-      "learning_rate": 0.00018999999999999998,
-      "loss": 2.6206,
+      "grad_norm": 0.8893711566925049,
+      "learning_rate": 0.0001425,
+      "loss": 3.292,
       "step": 481
     },
     {
       "epoch": 0.7712,
-      "grad_norm": 0.9115980863571167,
-      "learning_rate": 0.0001904,
-      "loss": 2.5279,
+      "grad_norm": 0.8357609510421753,
+      "learning_rate": 0.00014279999999999997,
+      "loss": 3.1505,
       "step": 482
     },
     {
       "epoch": 0.7728,
-      "grad_norm": 1.0715998411178589,
-      "learning_rate": 0.00019079999999999998,
-      "loss": 2.9977,
+      "grad_norm": 0.9251508712768555,
+      "learning_rate": 0.00014309999999999998,
+      "loss": 3.4092,
       "step": 483
     },
     {
       "epoch": 0.7744,
-      "grad_norm": 3.206315517425537,
-      "learning_rate": 0.00019119999999999999,
-      "loss": 2.7214,
+      "grad_norm": 4.293961524963379,
+      "learning_rate": 0.0001434,
+      "loss": 3.3033,
       "step": 484
     },
     {
       "epoch": 0.776,
-      "grad_norm": 1.8996236324310303,
-      "learning_rate": 0.0001916,
-      "loss": 2.5993,
+      "grad_norm": 3.240405559539795,
+      "learning_rate": 0.00014369999999999997,
+      "loss": 3.2177,
       "step": 485
     },
     {
       "epoch": 0.7776,
-      "grad_norm": 1.2291277647018433,
-      "learning_rate": 0.00019199999999999998,
-      "loss": 2.6469,
+      "grad_norm": 1.5192947387695312,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 3.2168,
       "step": 486
     },
     {
       "epoch": 0.7792,
-      "grad_norm": 1.7361717224121094,
-      "learning_rate": 0.0001924,
-      "loss": 2.7336,
+      "grad_norm": 0.9653074145317078,
+      "learning_rate": 0.00014429999999999998,
+      "loss": 3.2922,
       "step": 487
     },
     {
       "epoch": 0.7808,
-      "grad_norm": 1.7760653495788574,
-      "learning_rate": 0.0001928,
-      "loss": 2.7031,
+      "grad_norm": 4.4581475257873535,
+      "learning_rate": 0.0001446,
+      "loss": 3.2312,
       "step": 488
     },
     {
       "epoch": 0.7824,
-      "grad_norm": 2.499051094055176,
-      "learning_rate": 0.00019319999999999998,
-      "loss": 2.7771,
+      "grad_norm": 1.0327415466308594,
+      "learning_rate": 0.00014489999999999997,
+      "loss": 3.2242,
       "step": 489
     },
     {
       "epoch": 0.784,
-      "grad_norm": 1.7659521102905273,
-      "learning_rate": 0.0001936,
-      "loss": 2.6913,
+      "grad_norm": 1.4427021741867065,
+      "learning_rate": 0.00014519999999999998,
+      "loss": 3.2571,
       "step": 490
     },
     {
       "epoch": 0.7856,
-      "grad_norm": 1.1593416929244995,
-      "learning_rate": 0.00019399999999999997,
-      "loss": 2.8942,
+      "grad_norm": 1.1032732725143433,
+      "learning_rate": 0.00014549999999999999,
+      "loss": 3.3207,
       "step": 491
     },
     {
       "epoch": 0.7872,
-      "grad_norm": 1.008558988571167,
-      "learning_rate": 0.00019439999999999998,
-      "loss": 2.6859,
+      "grad_norm": 1.0624717473983765,
+      "learning_rate": 0.0001458,
+      "loss": 3.2055,
       "step": 492
     },
     {
       "epoch": 0.7888,
-      "grad_norm": 1.389395833015442,
-      "learning_rate": 0.0001948,
-      "loss": 2.4138,
+      "grad_norm": 1.9576382637023926,
+      "learning_rate": 0.00014609999999999997,
+      "loss": 3.0291,
       "step": 493
     },
     {
       "epoch": 0.7904,
-      "grad_norm": 1.649500846862793,
-      "learning_rate": 0.00019519999999999997,
-      "loss": 2.8378,
+      "grad_norm": 1.1364521980285645,
+      "learning_rate": 0.00014639999999999998,
+      "loss": 3.2193,
       "step": 494
     },
     {
       "epoch": 0.792,
-      "grad_norm": 1.2002882957458496,
-      "learning_rate": 0.00019559999999999998,
-      "loss": 2.5141,
+      "grad_norm": 1.5906617641448975,
+      "learning_rate": 0.0001467,
+      "loss": 3.2005,
       "step": 495
     },
     {
       "epoch": 0.7936,
-      "grad_norm": 2.0668468475341797,
-      "learning_rate": 0.00019599999999999997,
-      "loss": 2.4013,
+      "grad_norm": 2.4177134037017822,
+      "learning_rate": 0.000147,
+      "loss": 3.0561,
       "step": 496
     },
     {
       "epoch": 0.7952,
-      "grad_norm": 3.268202066421509,
-      "learning_rate": 0.00019639999999999998,
-      "loss": 2.664,
+      "grad_norm": 2.282595634460449,
+      "learning_rate": 0.00014729999999999998,
+      "loss": 3.2687,
       "step": 497
     },
     {
       "epoch": 0.7968,
-      "grad_norm": 1.8152614831924438,
-      "learning_rate": 0.00019679999999999999,
-      "loss": 2.6374,
+      "grad_norm": 3.5763862133026123,
+      "learning_rate": 0.00014759999999999998,
+      "loss": 3.1898,
       "step": 498
     },
     {
       "epoch": 0.7984,
-      "grad_norm": 2.2615838050842285,
-      "learning_rate": 0.00019719999999999997,
-      "loss": 2.9934,
+      "grad_norm": 3.573193311691284,
+      "learning_rate": 0.0001479,
+      "loss": 3.4237,
       "step": 499
     },
     {
       "epoch": 0.8,
-      "grad_norm": 2.3490400314331055,
-      "learning_rate": 0.00019759999999999998,
-      "loss": 3.0699,
-      "step": 500
-    },
-    {
-      "epoch": 0.8,
-      "eval_cer": 0.6001793757723124,
-      "eval_loss": 2.7479889392852783,
-      "eval_runtime": 158.7062,
-      "eval_samples_per_second": 19.76,
-      "eval_steps_per_second": 1.235,
-      "eval_wer": 0.8336959900528442,
+      "grad_norm": 3.274698495864868,
+      "learning_rate": 0.0001482,
+      "loss": 3.4449,
       "step": 500
     },
     {
       "epoch": 0.8016,
-      "grad_norm": 7.252814292907715,
-      "learning_rate": 0.000198,
-      "loss": 2.8085,
+      "grad_norm": 7.6723127365112305,
+      "learning_rate": 0.00014849999999999998,
+      "loss": 3.4171,
       "step": 501
     },
     {
       "epoch": 0.8032,
-      "grad_norm": 4.065989017486572,
-      "learning_rate": 0.00019839999999999997,
-      "loss": 2.5431,
+      "grad_norm": 3.750722646713257,
+      "learning_rate": 0.00014879999999999998,
+      "loss": 3.1539,
       "step": 502
     },
     {
       "epoch": 0.8048,
-      "grad_norm": 5.091382026672363,
-      "learning_rate": 0.00019879999999999998,
-      "loss": 2.8578,
+      "grad_norm": 6.335118770599365,
+      "learning_rate": 0.0001491,
+      "loss": 3.4048,
       "step": 503
     },
     {
       "epoch": 0.8064,
-      "grad_norm": 2.2774922847747803,
-      "learning_rate": 0.0001992,
-      "loss": 2.6625,
+      "grad_norm": 3.3972010612487793,
+      "learning_rate": 0.0001494,
+      "loss": 3.2258,
       "step": 504
     },
     {
       "epoch": 0.808,
-      "grad_norm": 2.239748239517212,
-      "learning_rate": 0.00019959999999999997,
-      "loss": 2.6375,
+      "grad_norm": 1.252905249595642,
+      "learning_rate": 0.00014969999999999998,
+      "loss": 3.2009,
       "step": 505
     },
     {
       "epoch": 0.8096,
-      "grad_norm": 3.0972511768341064,
-      "learning_rate": 0.00019999999999999998,
-      "loss": 2.9029,
+      "grad_norm": 1.956114411354065,
+      "learning_rate": 0.00015,
+      "loss": 3.3399,
       "step": 506
     },
     {
       "epoch": 0.8112,
-      "grad_norm": 5.276496410369873,
-      "learning_rate": 0.0002004,
-      "loss": 2.4195,
+      "grad_norm": 4.416980743408203,
+      "learning_rate": 0.0001503,
+      "loss": 3.0746,
       "step": 507
     },
     {
       "epoch": 0.8128,
-      "grad_norm": 4.866186618804932,
-      "learning_rate": 0.00020079999999999997,
-      "loss": 2.4582,
+      "grad_norm": 4.364480495452881,
+      "learning_rate": 0.00015059999999999997,
+      "loss": 3.0995,
       "step": 508
     },
     {
       "epoch": 0.8144,
-      "grad_norm": 3.0533981323242188,
-      "learning_rate": 0.00020119999999999998,
-      "loss": 2.4498,
+      "grad_norm": 3.2153031826019287,
+      "learning_rate": 0.00015089999999999998,
+      "loss": 3.1032,
       "step": 509
     },
     {
       "epoch": 0.816,
-      "grad_norm": 1.130332350730896,
-      "learning_rate": 0.0002016,
-      "loss": 2.4513,
+      "grad_norm": 1.0816490650177002,
+      "learning_rate": 0.0001512,
+      "loss": 3.0827,
       "step": 510
     },
     {
       "epoch": 0.8176,
-      "grad_norm": 1.124791145324707,
-      "learning_rate": 0.00020199999999999998,
-      "loss": 2.1947,
+      "grad_norm": 1.157222867012024,
+      "learning_rate": 0.0001515,
+      "loss": 2.9231,
       "step": 511
     },
     {
       "epoch": 0.8192,
-      "grad_norm": 3.729512929916382,
-      "learning_rate": 0.0002024,
-      "loss": 2.3357,
+      "grad_norm": 3.4473843574523926,
+      "learning_rate": 0.00015179999999999998,
+      "loss": 2.9558,
       "step": 512
     },
     {
       "epoch": 0.8208,
-      "grad_norm": 3.9868364334106445,
-      "learning_rate": 0.0002028,
-      "loss": 2.4641,
+      "grad_norm": 4.260528087615967,
+      "learning_rate": 0.00015209999999999998,
+      "loss": 3.0114,
       "step": 513
     },
     {
       "epoch": 0.8224,
-      "grad_norm": 3.2634570598602295,
-      "learning_rate": 0.00020319999999999998,
-      "loss": 2.2873,
+      "grad_norm": 3.8227248191833496,
+      "learning_rate": 0.0001524,
+      "loss": 2.912,
       "step": 514
     },
     {
       "epoch": 0.824,
-      "grad_norm": 6.407214641571045,
-      "learning_rate": 0.00020359999999999996,
-      "loss": 2.4176,
+      "grad_norm": 7.646665573120117,
+      "learning_rate": 0.0001527,
+      "loss": 3.1023,
       "step": 515
     },
     {
       "epoch": 0.8256,
-      "grad_norm": 2.0617222785949707,
-      "learning_rate": 0.000204,
-      "loss": 2.3114,
+      "grad_norm": 2.7784230709075928,
+      "learning_rate": 0.00015299999999999998,
+      "loss": 2.9334,
       "step": 516
     },
     {
       "epoch": 0.8272,
-      "grad_norm": 1.7285425662994385,
-      "learning_rate": 0.00020439999999999998,
-      "loss": 2.2132,
+      "grad_norm": 2.957230806350708,
+      "learning_rate": 0.00015329999999999999,
+      "loss": 2.9305,
       "step": 517
     },
     {
       "epoch": 0.8288,
-      "grad_norm": 1.2327693700790405,
-      "learning_rate": 0.00020479999999999996,
-      "loss": 1.9775,
+      "grad_norm": 1.1709569692611694,
+      "learning_rate": 0.0001536,
+      "loss": 2.7356,
       "step": 518
     },
     {
       "epoch": 0.8304,
-      "grad_norm": 1.495557427406311,
-      "learning_rate": 0.0002052,
-      "loss": 2.3095,
+      "grad_norm": 1.495569109916687,
+      "learning_rate": 0.0001539,
+      "loss": 2.9082,
       "step": 519
     },
     {
       "epoch": 0.832,
-      "grad_norm": 1.9202693700790405,
-      "learning_rate": 0.00020559999999999998,
-      "loss": 2.3414,
+      "grad_norm": 2.442965269088745,
+      "learning_rate": 0.00015419999999999998,
+      "loss": 2.8912,
       "step": 520
     },
     {
       "epoch": 0.8336,
-      "grad_norm": 1.3025583028793335,
-      "learning_rate": 0.00020599999999999997,
-      "loss": 2.4883,
+      "grad_norm": 1.7007681131362915,
+      "learning_rate": 0.0001545,
+      "loss": 2.9658,
       "step": 521
     },
     {
       "epoch": 0.8352,
-      "grad_norm": 1.6052091121673584,
-      "learning_rate": 0.00020639999999999998,
-      "loss": 2.1605,
+      "grad_norm": 2.2086222171783447,
+      "learning_rate": 0.0001548,
+      "loss": 2.7858,
       "step": 522
     },
     {
       "epoch": 0.8368,
-      "grad_norm": 1.410342812538147,
-      "learning_rate": 0.00020679999999999999,
-      "loss": 2.2074,
+      "grad_norm": 1.53826904296875,
+      "learning_rate": 0.0001551,
+      "loss": 2.7899,
       "step": 523
     },
     {
       "epoch": 0.8384,
-      "grad_norm": 2.4669413566589355,
-      "learning_rate": 0.00020719999999999997,
-      "loss": 2.6925,
+      "grad_norm": 2.381734848022461,
+      "learning_rate": 0.00015539999999999998,
+      "loss": 3.1133,
       "step": 524
     },
     {
       "epoch": 0.84,
-      "grad_norm": 2.8002283573150635,
-      "learning_rate": 0.00020759999999999998,
-      "loss": 2.4361,
+      "grad_norm": 3.017554998397827,
+      "learning_rate": 0.0001557,
+      "loss": 2.9739,
       "step": 525
     },
     {
       "epoch": 0.8416,
-      "grad_norm": 2.760272264480591,
-      "learning_rate": 0.000208,
-      "loss": 2.3288,
+      "grad_norm": 3.71621036529541,
+      "learning_rate": 0.000156,
+      "loss": 2.8974,
       "step": 526
     },
     {
       "epoch": 0.8432,
-      "grad_norm": 1.630889654159546,
-      "learning_rate": 0.00020839999999999997,
-      "loss": 2.2201,
+      "grad_norm": 1.7555559873580933,
+      "learning_rate": 0.0001563,
+      "loss": 2.8068,
       "step": 527
     },
     {
       "epoch": 0.8448,
-      "grad_norm": 1.1679999828338623,
-      "learning_rate": 0.00020879999999999998,
-      "loss": 2.1824,
+      "grad_norm": 1.1658614873886108,
+      "learning_rate": 0.00015659999999999998,
+      "loss": 2.7185,
       "step": 528
     },
     {
       "epoch": 0.8464,
-      "grad_norm": 1.2356942892074585,
-      "learning_rate": 0.0002092,
-      "loss": 2.0637,
+      "grad_norm": 0.8518783450126648,
+      "learning_rate": 0.0001569,
+      "loss": 2.7038,
       "step": 529
     },
     {
       "epoch": 0.848,
-      "grad_norm": 0.9838883876800537,
-      "learning_rate": 0.00020959999999999997,
-      "loss": 2.1315,
+      "grad_norm": 1.1382691860198975,
+      "learning_rate": 0.0001572,
+      "loss": 2.7108,
       "step": 530
     },
     {
       "epoch": 0.8496,
-      "grad_norm": 2.134653091430664,
-      "learning_rate": 0.00020999999999999998,
-      "loss": 2.3373,
+      "grad_norm": 1.3236608505249023,
+      "learning_rate": 0.00015749999999999998,
+      "loss": 2.862,
       "step": 531
     },
     {
       "epoch": 0.8512,
-      "grad_norm": 1.8091658353805542,
-      "learning_rate": 0.0002104,
-      "loss": 2.1348,
+      "grad_norm": 1.4057608842849731,
+      "learning_rate": 0.0001578,
+      "loss": 2.7198,
       "step": 532
     },
     {
       "epoch": 0.8528,
-      "grad_norm": 1.6032664775848389,
-      "learning_rate": 0.00021079999999999997,
-      "loss": 1.9647,
+      "grad_norm": 1.2580957412719727,
+      "learning_rate": 0.0001581,
+      "loss": 2.5294,
       "step": 533
     },
     {
       "epoch": 0.8544,
-      "grad_norm": 1.6092320680618286,
-      "learning_rate": 0.00021119999999999996,
-      "loss": 2.1598,
+      "grad_norm": 1.328491449356079,
+      "learning_rate": 0.0001584,
+      "loss": 2.6866,
       "step": 534
     },
     {
       "epoch": 0.856,
-      "grad_norm": 1.3249115943908691,
-      "learning_rate": 0.0002116,
-      "loss": 2.4795,
+      "grad_norm": 1.0519988536834717,
+      "learning_rate": 0.00015869999999999998,
+      "loss": 2.8501,
       "step": 535
     },
     {
       "epoch": 0.8576,
-      "grad_norm": 1.057600975036621,
-      "learning_rate": 0.00021199999999999998,
-      "loss": 2.5951,
+      "grad_norm": 1.733718752861023,
+      "learning_rate": 0.000159,
+      "loss": 2.9786,
       "step": 536
     },
     {
       "epoch": 0.8592,
-      "grad_norm": 1.3457427024841309,
-      "learning_rate": 0.00021239999999999996,
-      "loss": 1.9831,
+      "grad_norm": 1.1301422119140625,
+      "learning_rate": 0.0001593,
+      "loss": 2.5756,
       "step": 537
     },
     {
       "epoch": 0.8608,
-      "grad_norm": 1.3885570764541626,
-      "learning_rate": 0.0002128,
-      "loss": 1.9622,
+      "grad_norm": 1.241924524307251,
+      "learning_rate": 0.0001596,
+      "loss": 2.5149,
       "step": 538
     },
     {
       "epoch": 0.8624,
-      "grad_norm": 1.6927804946899414,
-      "learning_rate": 0.00021319999999999998,
-      "loss": 2.2762,
+      "grad_norm": 1.328609824180603,
+      "learning_rate": 0.00015989999999999998,
+      "loss": 2.7648,
       "step": 539
     },
     {
       "epoch": 0.864,
-      "grad_norm": 1.5378133058547974,
-      "learning_rate": 0.00021359999999999996,
-      "loss": 2.1121,
+      "grad_norm": 1.236745834350586,
+      "learning_rate": 0.0001602,
+      "loss": 2.5732,
       "step": 540
     },
     {
       "epoch": 0.8656,
-      "grad_norm": 2.0830371379852295,
-      "learning_rate": 0.000214,
-      "loss": 2.0769,
+      "grad_norm": 1.5136185884475708,
+      "learning_rate": 0.0001605,
+      "loss": 2.5341,
       "step": 541
     },
     {
       "epoch": 0.8672,
-      "grad_norm": 1.5723661184310913,
-      "learning_rate": 0.00021439999999999998,
-      "loss": 2.48,
+      "grad_norm": 1.1722863912582397,
+      "learning_rate": 0.0001608,
+      "loss": 2.9393,
       "step": 542
     },
     {
       "epoch": 0.8688,
-      "grad_norm": 2.520364999771118,
-      "learning_rate": 0.00021479999999999996,
-      "loss": 2.293,
+      "grad_norm": 1.9231904745101929,
+      "learning_rate": 0.00016109999999999999,
+      "loss": 2.674,
       "step": 543
     },
     {
       "epoch": 0.8704,
-      "grad_norm": 1.5732394456863403,
-      "learning_rate": 0.0002152,
-      "loss": 1.9714,
+      "grad_norm": 1.632614254951477,
+      "learning_rate": 0.0001614,
+      "loss": 2.5229,
       "step": 544
     },
     {
       "epoch": 0.872,
-      "grad_norm": 1.117932677268982,
-      "learning_rate": 0.00021559999999999998,
-      "loss": 2.267,
+      "grad_norm": 1.2549848556518555,
+      "learning_rate": 0.0001617,
+      "loss": 2.6451,
       "step": 545
     },
     {
       "epoch": 0.8736,
       "grad_norm": NaN,
-      "learning_rate": 0.00021559999999999998,
-      "loss": 2.4488,
+      "learning_rate": 0.0001617,
+      "loss": 2.7689,
       "step": 546
     },
     {
       "epoch": 0.8752,
-      "grad_norm": 2.9612374305725098,
-      "learning_rate": 0.00021599999999999996,
-      "loss": 2.674,
+      "grad_norm": 1.994581937789917,
+      "learning_rate": 0.000162,
+      "loss": 2.9942,
       "step": 547
     },
     {
       "epoch": 0.8768,
-      "grad_norm": 4.073216915130615,
-      "learning_rate": 0.0002164,
-      "loss": 2.7163,
+      "grad_norm": 2.948986291885376,
+      "learning_rate": 0.0001623,
+      "loss": 2.9567,
       "step": 548
     },
     {
       "epoch": 0.8784,
-      "grad_norm": 1.3353551626205444,
-      "learning_rate": 0.00021679999999999998,
-      "loss": 2.4723,
+      "grad_norm": 3.3447182178497314,
+      "learning_rate": 0.0001626,
+      "loss": 2.926,
       "step": 549
     },
     {
       "epoch": 0.88,
-      "grad_norm": 2.927354097366333,
-      "learning_rate": 0.00021719999999999997,
-      "loss": 2.6158,
+      "grad_norm": 1.6422103643417358,
+      "learning_rate": 0.0001629,
+      "loss": 2.9978,
       "step": 550
     },
     {
       "epoch": 0.8816,
-      "grad_norm": 7.480506420135498,
-      "learning_rate": 0.0002176,
-      "loss": 2.6206,
+      "grad_norm": 17.27240753173828,
+      "learning_rate": 0.0001632,
+      "loss": 3.4376,
       "step": 551
     },
     {
       "epoch": 0.8832,
-      "grad_norm": 10.748481750488281,
-      "learning_rate": 0.00021799999999999999,
-      "loss": 3.1361,
+      "grad_norm": 16.129791259765625,
+      "learning_rate": 0.0001635,
+      "loss": 3.5866,
       "step": 552
     },
     {
       "epoch": 0.8848,
-      "grad_norm": 4.278016090393066,
-      "learning_rate": 0.00021839999999999997,
-      "loss": 2.3451,
+      "grad_norm": 6.709097862243652,
+      "learning_rate": 0.0001638,
+      "loss": 2.8288,
       "step": 553
     },
     {
       "epoch": 0.8864,
-      "grad_norm": 11.78780460357666,
-      "learning_rate": 0.00021879999999999995,
-      "loss": 3.1178,
+      "grad_norm": 15.814329147338867,
+      "learning_rate": 0.0001641,
+      "loss": 3.5662,
       "step": 554
     },
     {
       "epoch": 0.888,
-      "grad_norm": 2.092212438583374,
-      "learning_rate": 0.0002192,
-      "loss": 2.3741,
+      "grad_norm": 1.16456937789917,
+      "learning_rate": 0.0001644,
+      "loss": 2.7902,
       "step": 555
     },
     {
       "epoch": 0.8896,
-      "grad_norm": 2.331865072250366,
-      "learning_rate": 0.00021959999999999997,
-      "loss": 1.9955,
+      "grad_norm": 1.7065085172653198,
+      "learning_rate": 0.0001647,
+      "loss": 2.4504,
       "step": 556
     },
     {
       "epoch": 0.8912,
-      "grad_norm": 2.80426025390625,
-      "learning_rate": 0.00021999999999999995,
-      "loss": 2.0726,
+      "grad_norm": 2.260443925857544,
+      "learning_rate": 0.000165,
+      "loss": 2.5193,
       "step": 557
     },
     {
       "epoch": 0.8928,
-      "grad_norm": 2.815202474594116,
-      "learning_rate": 0.0002204,
-      "loss": 2.4074,
+      "grad_norm": 2.0311453342437744,
+      "learning_rate": 0.0001653,
+      "loss": 2.7037,
       "step": 558
     },
     {
       "epoch": 0.8944,
-      "grad_norm": 2.3765721321105957,
-      "learning_rate": 0.00022079999999999997,
-      "loss": 2.206,
+      "grad_norm": 2.1808862686157227,
+      "learning_rate": 0.0001656,
+      "loss": 2.5959,
       "step": 559
     },
     {
       "epoch": 0.896,
-      "grad_norm": 6.6429314613342285,
-      "learning_rate": 0.00022119999999999996,
-      "loss": 2.4754,
+      "grad_norm": 6.377316951751709,
+      "learning_rate": 0.0001659,
+      "loss": 2.839,
       "step": 560
     },
     {
       "epoch": 0.8976,
-      "grad_norm": 5.93609094619751,
-      "learning_rate": 0.0002216,
-      "loss": 2.7358,
+      "grad_norm": 6.089463710784912,
+      "learning_rate": 0.0001662,
+      "loss": 3.1665,
       "step": 561
     },
     {
       "epoch": 0.8992,
-      "grad_norm": 6.85822868347168,
-      "learning_rate": 0.00022199999999999998,
-      "loss": 2.459,
+      "grad_norm": 6.491182804107666,
+      "learning_rate": 0.0001665,
+      "loss": 2.885,
       "step": 562
     },
     {
       "epoch": 0.9008,
-      "grad_norm": 0.9314383864402771,
-      "learning_rate": 0.00022239999999999996,
-      "loss": 1.8933,
+      "grad_norm": 1.3139764070510864,
+      "learning_rate": 0.0001668,
+      "loss": 2.3777,
       "step": 563
     },
     {
       "epoch": 0.9024,
-      "grad_norm": 1.0340198278427124,
-      "learning_rate": 0.0002228,
-      "loss": 2.0382,
+      "grad_norm": 1.7619560956954956,
+      "learning_rate": 0.0001671,
+      "loss": 2.5143,
       "step": 564
     },
     {
       "epoch": 0.904,
-      "grad_norm": 1.108193278312683,
-      "learning_rate": 0.00022319999999999998,
-      "loss": 2.0965,
+      "grad_norm": 1.494608759880066,
+      "learning_rate": 0.0001674,
+      "loss": 2.4989,
       "step": 565
     },
     {
       "epoch": 0.9056,
-      "grad_norm": 1.070430040359497,
-      "learning_rate": 0.00022359999999999996,
-      "loss": 1.9219,
+      "grad_norm": 1.0623366832733154,
+      "learning_rate": 0.0001677,
+      "loss": 2.3889,
       "step": 566
     },
     {
       "epoch": 0.9072,
-      "grad_norm": 1.2231314182281494,
-      "learning_rate": 0.000224,
-      "loss": 1.8723,
+      "grad_norm": 1.6987491846084595,
+      "learning_rate": 0.000168,
+      "loss": 2.3293,
       "step": 567
     },
     {
       "epoch": 0.9088,
-      "grad_norm": 1.2423832416534424,
-      "learning_rate": 0.00022439999999999998,
-      "loss": 1.7471,
+      "grad_norm": 2.076875925064087,
+      "learning_rate": 0.0001683,
+      "loss": 2.2478,
       "step": 568
     },
     {
       "epoch": 0.9104,
-      "grad_norm": 1.246390461921692,
-      "learning_rate": 0.00022479999999999996,
-      "loss": 1.8937,
+      "grad_norm": 2.2903881072998047,
+      "learning_rate": 0.0001686,
+      "loss": 2.3652,
       "step": 569
     },
     {
       "epoch": 0.912,
-      "grad_norm": 1.130929946899414,
-      "learning_rate": 0.0002252,
-      "loss": 1.8579,
+      "grad_norm": 2.357800245285034,
+      "learning_rate": 0.00016889999999999996,
+      "loss": 2.298,
       "step": 570
     },
     {
       "epoch": 0.9136,
-      "grad_norm": 1.304247260093689,
-      "learning_rate": 0.00022559999999999998,
-      "loss": 2.0576,
+      "grad_norm": 0.7803952693939209,
+      "learning_rate": 0.00016919999999999997,
+      "loss": 2.4371,
       "step": 571
     },
     {
       "epoch": 0.9152,
-      "grad_norm": 1.309567928314209,
-      "learning_rate": 0.00022599999999999996,
-      "loss": 1.9817,
+      "grad_norm": 1.7425607442855835,
+      "learning_rate": 0.00016949999999999997,
+      "loss": 2.3749,
       "step": 572
     },
     {
       "epoch": 0.9168,
-      "grad_norm": 1.2422513961791992,
-      "learning_rate": 0.0002264,
-      "loss": 2.2693,
+      "grad_norm": 1.223272681236267,
+      "learning_rate": 0.00016979999999999998,
+      "loss": 2.5863,
       "step": 573
     },
     {
       "epoch": 0.9184,
-      "grad_norm": 1.0724761486053467,
-      "learning_rate": 0.00022679999999999998,
-      "loss": 1.9313,
+      "grad_norm": 1.270267128944397,
+      "learning_rate": 0.00017009999999999996,
+      "loss": 2.3319,
       "step": 574
     },
     {
       "epoch": 0.92,
-      "grad_norm": 1.7489149570465088,
-      "learning_rate": 0.00022719999999999997,
-      "loss": 2.1692,
+      "grad_norm": 1.346740961074829,
+      "learning_rate": 0.00017039999999999997,
+      "loss": 2.5017,
       "step": 575
     },
     {
       "epoch": 0.9216,
-      "grad_norm": 1.168873906135559,
-      "learning_rate": 0.0002276,
-      "loss": 2.0218,
+      "grad_norm": 1.2833489179611206,
+      "learning_rate": 0.00017069999999999998,
+      "loss": 2.374,
       "step": 576
     },
     {
       "epoch": 0.9232,
-      "grad_norm": 0.9614212512969971,
-      "learning_rate": 0.00022799999999999999,
-      "loss": 2.3924,
+      "grad_norm": 1.1349873542785645,
+      "learning_rate": 0.00017099999999999998,
+      "loss": 2.7307,
       "step": 577
     },
     {
       "epoch": 0.9248,
-      "grad_norm": 1.258357048034668,
-      "learning_rate": 0.00022839999999999997,
-      "loss": 1.9544,
+      "grad_norm": 1.443171501159668,
+      "learning_rate": 0.00017129999999999996,
+      "loss": 2.3192,
       "step": 578
     },
     {
       "epoch": 0.9264,
-      "grad_norm": 2.0249435901641846,
-      "learning_rate": 0.0002288,
-      "loss": 1.9929,
+      "grad_norm": 1.5835188627243042,
+      "learning_rate": 0.00017159999999999997,
+      "loss": 2.3311,
       "step": 579
     },
     {
       "epoch": 0.928,
-      "grad_norm": 1.0736762285232544,
-      "learning_rate": 0.0002292,
-      "loss": 1.7423,
+      "grad_norm": 1.2339696884155273,
+      "learning_rate": 0.00017189999999999998,
+      "loss": 2.1611,
       "step": 580
     },
     {
       "epoch": 0.9296,
-      "grad_norm": 1.4606719017028809,
-      "learning_rate": 0.00022959999999999997,
-      "loss": 1.8804,
+      "grad_norm": 1.6790226697921753,
+      "learning_rate": 0.00017219999999999998,
+      "loss": 2.2807,
       "step": 581
     },
     {
       "epoch": 0.9312,
-      "grad_norm": 0.8972280025482178,
-      "learning_rate": 0.00023,
-      "loss": 1.8398,
+      "grad_norm": 1.297339677810669,
+      "learning_rate": 0.00017249999999999996,
+      "loss": 2.2204,
       "step": 582
     },
     {
       "epoch": 0.9328,
-      "grad_norm": 2.3839757442474365,
-      "learning_rate": 0.0002304,
-      "loss": 2.1795,
+      "grad_norm": 2.8306925296783447,
+      "learning_rate": 0.00017279999999999997,
+      "loss": 2.4957,
       "step": 583
     },
     {
       "epoch": 0.9344,
-      "grad_norm": 1.3923808336257935,
-      "learning_rate": 0.00023079999999999997,
-      "loss": 1.7968,
+      "grad_norm": 1.6782737970352173,
+      "learning_rate": 0.00017309999999999998,
+      "loss": 2.1674,
       "step": 584
     },
     {
       "epoch": 0.936,
-      "grad_norm": 0.9030740857124329,
-      "learning_rate": 0.0002312,
-      "loss": 1.6572,
+      "grad_norm": 0.9002209305763245,
+      "learning_rate": 0.00017339999999999996,
+      "loss": 2.0584,
       "step": 585
     },
     {
       "epoch": 0.9376,
-      "grad_norm": 1.322900414466858,
-      "learning_rate": 0.0002316,
-      "loss": 1.8275,
+      "grad_norm": 1.204094409942627,
+      "learning_rate": 0.00017369999999999997,
+      "loss": 2.1932,
       "step": 586
     },
     {
       "epoch": 0.9392,
-      "grad_norm": 1.1124223470687866,
-      "learning_rate": 0.00023199999999999997,
-      "loss": 2.0224,
+      "grad_norm": 1.075980544090271,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 2.3717,
       "step": 587
     },
     {
       "epoch": 0.9408,
-      "grad_norm": 2.4704806804656982,
-      "learning_rate": 0.00023239999999999996,
-      "loss": 1.9902,
+      "grad_norm": 1.9965050220489502,
+      "learning_rate": 0.00017429999999999998,
+      "loss": 2.2855,
       "step": 588
     },
     {
       "epoch": 0.9424,
-      "grad_norm": 2.128262758255005,
-      "learning_rate": 0.0002328,
-      "loss": 1.795,
+      "grad_norm": 2.01662015914917,
+      "learning_rate": 0.00017459999999999996,
+      "loss": 2.1505,
       "step": 589
     },
     {
       "epoch": 0.944,
-      "grad_norm": 1.577215313911438,
-      "learning_rate": 0.00023319999999999998,
-      "loss": 2.0014,
+      "grad_norm": 1.4860492944717407,
+      "learning_rate": 0.00017489999999999997,
+      "loss": 2.2968,
       "step": 590
     },
     {
       "epoch": 0.9456,
-      "grad_norm": 1.1099101305007935,
-      "learning_rate": 0.00023359999999999996,
-      "loss": 1.9604,
+      "grad_norm": 1.2854816913604736,
+      "learning_rate": 0.00017519999999999998,
+      "loss": 2.3285,
       "step": 591
     },
     {
       "epoch": 0.9472,
-      "grad_norm": 1.6668699979782104,
-      "learning_rate": 0.000234,
-      "loss": 2.2408,
+      "grad_norm": 1.5911052227020264,
+      "learning_rate": 0.00017549999999999998,
+      "loss": 2.3975,
       "step": 592
     },
     {
       "epoch": 0.9488,
-      "grad_norm": 1.8698573112487793,
-      "learning_rate": 0.00023439999999999998,
-      "loss": 2.0816,
+      "grad_norm": 1.413689374923706,
+      "learning_rate": 0.00017579999999999996,
+      "loss": 2.3824,
       "step": 593
     },
     {
       "epoch": 0.9504,
-      "grad_norm": 2.9194648265838623,
-      "learning_rate": 0.00023479999999999996,
-      "loss": 1.9059,
+      "grad_norm": 2.4783072471618652,
+      "learning_rate": 0.00017609999999999997,
+      "loss": 2.1916,
       "step": 594
     },
     {
       "epoch": 0.952,
-      "grad_norm": 1.137349009513855,
-      "learning_rate": 0.0002352,
-      "loss": 2.0177,
+      "grad_norm": 1.7629774808883667,
+      "learning_rate": 0.00017639999999999998,
+      "loss": 2.33,
       "step": 595
     },
     {
       "epoch": 0.9536,
-      "grad_norm": 1.3729033470153809,
-      "learning_rate": 0.00023559999999999998,
-      "loss": 2.0672,
+      "grad_norm": 1.8146947622299194,
+      "learning_rate": 0.00017669999999999999,
+      "loss": 2.4165,
       "step": 596
     },
     {
       "epoch": 0.9552,
-      "grad_norm": 1.4698584079742432,
-      "learning_rate": 0.00023599999999999996,
-      "loss": 1.8058,
+      "grad_norm": 1.5814777612686157,
+      "learning_rate": 0.00017699999999999997,
+      "loss": 2.1172,
       "step": 597
     },
     {
       "epoch": 0.9568,
-      "grad_norm": 1.9527052640914917,
-      "learning_rate": 0.0002364,
-      "loss": 2.1384,
+      "grad_norm": 1.7164720296859741,
+      "learning_rate": 0.00017729999999999997,
+      "loss": 2.3839,
       "step": 598
     },
     {
       "epoch": 0.9584,
-      "grad_norm": 1.9304373264312744,
-      "learning_rate": 0.00023679999999999998,
-      "loss": 2.0941,
+      "grad_norm": 1.7895351648330688,
+      "learning_rate": 0.00017759999999999998,
+      "loss": 2.4406,
       "step": 599
     },
     {
       "epoch": 0.96,
       "grad_norm": NaN,
-      "learning_rate": 0.00023679999999999998,
-      "loss": 1.7389,
-      "step": 600
-    },
-    {
-      "epoch": 0.96,
-      "eval_cer": 0.47166660023119544,
-      "eval_loss": 2.4468393325805664,
-      "eval_runtime": 158.3967,
-      "eval_samples_per_second": 19.798,
-      "eval_steps_per_second": 1.237,
-      "eval_wer": 0.6782302352087867,
+      "learning_rate": 0.00017759999999999998,
+      "loss": 2.1268,
       "step": 600
     },
     {
       "epoch": 0.9616,
-      "grad_norm": 22.909198760986328,
-      "learning_rate": 0.00023719999999999997,
-      "loss": 4.3211,
+      "grad_norm": 28.912076950073242,
+      "learning_rate": 0.0001779,
+      "loss": 4.5621,
       "step": 601
     },
     {
       "epoch": 0.9632,
-      "grad_norm": 6.32504415512085,
-      "learning_rate": 0.0002376,
-      "loss": 2.4554,
+      "grad_norm": 9.135653495788574,
+      "learning_rate": 0.00017819999999999997,
+      "loss": 2.7774,
       "step": 602
     },
     {
       "epoch": 0.9648,
-      "grad_norm": 2.5991179943084717,
-      "learning_rate": 0.00023799999999999998,
-      "loss": 2.2983,
+      "grad_norm": 4.935175895690918,
+      "learning_rate": 0.00017849999999999997,
+      "loss": 2.6115,
       "step": 603
     },
     {
       "epoch": 0.9664,
-      "grad_norm": 2.1994707584381104,
-      "learning_rate": 0.00023839999999999997,
-      "loss": 1.9377,
+      "grad_norm": 1.8409152030944824,
+      "learning_rate": 0.00017879999999999998,
+      "loss": 2.1965,
       "step": 604
     },
     {
       "epoch": 0.968,
-      "grad_norm": 3.0079381465911865,
-      "learning_rate": 0.0002388,
-      "loss": 2.0362,
+      "grad_norm": 1.7993711233139038,
+      "learning_rate": 0.0001791,
+      "loss": 2.2617,
       "step": 605
     },
     {
       "epoch": 0.9696,
-      "grad_norm": 2.14380145072937,
-      "learning_rate": 0.0002392,
-      "loss": 1.761,
+      "grad_norm": 1.4365873336791992,
+      "learning_rate": 0.00017939999999999997,
+      "loss": 2.0682,
       "step": 606
     },
     {
       "epoch": 0.9712,
-      "grad_norm": 3.1852333545684814,
-      "learning_rate": 0.00023959999999999997,
-      "loss": 1.8808,
+      "grad_norm": 3.2426984310150146,
+      "learning_rate": 0.00017969999999999998,
+      "loss": 2.124,
       "step": 607
     },
     {
       "epoch": 0.9728,
-      "grad_norm": 1.4008501768112183,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 1.9098,
+      "grad_norm": 1.8241300582885742,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 2.204,
       "step": 608
     },
     {
       "epoch": 0.9744,
-      "grad_norm": 1.1618804931640625,
-      "learning_rate": 0.0002404,
-      "loss": 1.95,
+      "grad_norm": 1.8929377794265747,
+      "learning_rate": 0.00018029999999999996,
+      "loss": 2.194,
       "step": 609
     },
     {
       "epoch": 0.976,
-      "grad_norm": 3.1721315383911133,
-      "learning_rate": 0.00024079999999999997,
-      "loss": 1.9567,
+      "grad_norm": 1.9729321002960205,
+      "learning_rate": 0.00018059999999999997,
+      "loss": 2.1454,
       "step": 610
     },
     {
       "epoch": 0.9776,
-      "grad_norm": 3.0783185958862305,
-      "learning_rate": 0.00024119999999999998,
-      "loss": 1.7904,
+      "grad_norm": 2.0208892822265625,
+      "learning_rate": 0.00018089999999999998,
+      "loss": 2.0107,
       "step": 611
     },
     {
       "epoch": 0.9792,
-      "grad_norm": 2.50898814201355,
-      "learning_rate": 0.0002416,
-      "loss": 1.8343,
+      "grad_norm": 2.0599329471588135,
+      "learning_rate": 0.00018119999999999999,
+      "loss": 2.1194,
       "step": 612
     },
     {
       "epoch": 0.9808,
-      "grad_norm": 1.6948851346969604,
-      "learning_rate": 0.00024199999999999997,
-      "loss": 1.7023,
+      "grad_norm": 1.338659405708313,
+      "learning_rate": 0.00018149999999999997,
+      "loss": 1.9717,
       "step": 613
     },
     {
       "epoch": 0.9824,
-      "grad_norm": 1.2017195224761963,
-      "learning_rate": 0.00024239999999999998,
-      "loss": 1.8026,
+      "grad_norm": 1.0782694816589355,
+      "learning_rate": 0.00018179999999999997,
+      "loss": 2.0979,
       "step": 614
     },
     {
       "epoch": 0.984,
-      "grad_norm": 1.4887975454330444,
-      "learning_rate": 0.0002428,
-      "loss": 1.6911,
+      "grad_norm": 1.9497066736221313,
+      "learning_rate": 0.00018209999999999998,
+      "loss": 1.9913,
       "step": 615
     },
     {
       "epoch": 0.9856,
-      "grad_norm": 0.9019091725349426,
-      "learning_rate": 0.00024319999999999998,
-      "loss": 1.5794,
+      "grad_norm": 1.055597186088562,
+      "learning_rate": 0.0001824,
+      "loss": 1.8407,
       "step": 616
     },
     {
       "epoch": 0.9872,
-      "grad_norm": 1.0633786916732788,
-      "learning_rate": 0.00024359999999999999,
-      "loss": 1.9869,
+      "grad_norm": 1.0797630548477173,
+      "learning_rate": 0.00018269999999999997,
+      "loss": 2.2538,
       "step": 617
     },
     {
       "epoch": 0.9888,
-      "grad_norm": 2.547787666320801,
-      "learning_rate": 0.000244,
-      "loss": 2.0831,
+      "grad_norm": 2.643570899963379,
+      "learning_rate": 0.00018299999999999998,
+      "loss": 2.3391,
       "step": 618
     },
     {
       "epoch": 0.9904,
-      "grad_norm": 2.7645835876464844,
-      "learning_rate": 0.0002444,
-      "loss": 1.8612,
+      "grad_norm": 2.8924105167388916,
+      "learning_rate": 0.00018329999999999998,
+      "loss": 2.1194,
       "step": 619
     },
     {
       "epoch": 0.992,
-      "grad_norm": 2.024125814437866,
-      "learning_rate": 0.0002448,
-      "loss": 1.683,
+      "grad_norm": 2.5089101791381836,
+      "learning_rate": 0.0001836,
+      "loss": 1.9541,
       "step": 620
     },
     {
       "epoch": 0.9936,
-      "grad_norm": 1.5566396713256836,
-      "learning_rate": 0.0002452,
-      "loss": 2.254,
+      "grad_norm": 1.9143519401550293,
+      "learning_rate": 0.00018389999999999997,
+      "loss": 2.5036,
       "step": 621
     },
     {
       "epoch": 0.9952,
-      "grad_norm": 1.9360650777816772,
-      "learning_rate": 0.00024559999999999995,
-      "loss": 1.9255,
+      "grad_norm": 2.4039244651794434,
+      "learning_rate": 0.00018419999999999998,
+      "loss": 2.2187,
       "step": 622
     },
     {
       "epoch": 0.9968,
-      "grad_norm": 1.7478992938995361,
-      "learning_rate": 0.00024599999999999996,
-      "loss": 2.09,
+      "grad_norm": 1.4363809823989868,
+      "learning_rate": 0.00018449999999999999,
+      "loss": 2.3675,
       "step": 623
     },
     {
       "epoch": 0.9984,
-      "grad_norm": 1.9008229970932007,
-      "learning_rate": 0.00024639999999999997,
-      "loss": 1.9071,
+      "grad_norm": 1.517674446105957,
+      "learning_rate": 0.0001848,
+      "loss": 2.0923,
       "step": 624
     },
     {
       "epoch": 1.0,
-      "grad_norm": 4.15518856048584,
-      "learning_rate": 0.0002468,
-      "loss": 2.1594,
+      "grad_norm": 3.559779167175293,
+      "learning_rate": 0.00018509999999999997,
+      "loss": 2.3816,
       "step": 625
     },
     {
       "epoch": 1.0016,
-      "grad_norm": 23.750652313232422,
-      "learning_rate": 0.0002472,
-      "loss": 3.5965,
+      "grad_norm": 23.562456130981445,
+      "learning_rate": 0.00018539999999999998,
+      "loss": 4.1691,
       "step": 626
     },
     {
       "epoch": 1.0032,
-      "grad_norm": 11.189056396484375,
-      "learning_rate": 0.0002476,
-      "loss": 2.7926,
+      "grad_norm": 8.592154502868652,
+      "learning_rate": 0.0001857,
+      "loss": 2.4914,
       "step": 627
     },
     {
       "epoch": 1.0048,
-      "grad_norm": 12.417805671691895,
-      "learning_rate": 0.00024799999999999996,
-      "loss": 3.3245,
+      "grad_norm": 6.448266506195068,
+      "learning_rate": 0.000186,
+      "loss": 2.2136,
       "step": 628
     },
     {
       "epoch": 1.0064,
-      "grad_norm": 4.926300525665283,
-      "learning_rate": 0.00024839999999999997,
-      "loss": 2.2806,
+      "grad_norm": 5.093257427215576,
+      "learning_rate": 0.00018629999999999997,
+      "loss": 2.831,
       "step": 629
     },
     {
       "epoch": 1.008,
-      "grad_norm": 1.0334339141845703,
-      "learning_rate": 0.0002488,
-      "loss": 2.0561,
+      "grad_norm": 1.622463345527649,
+      "learning_rate": 0.00018659999999999998,
+      "loss": 2.5009,
       "step": 630
     },
     {
       "epoch": 1.0096,
-      "grad_norm": 1.211844801902771,
-      "learning_rate": 0.0002492,
-      "loss": 1.9468,
+      "grad_norm": 2.0858829021453857,
+      "learning_rate": 0.0001869,
+      "loss": 2.3197,
       "step": 631
     },
     {
       "epoch": 1.0112,
-      "grad_norm": 5.117887496948242,
-      "learning_rate": 0.00024959999999999994,
-      "loss": 2.0291,
+      "grad_norm": 5.908426284790039,
+      "learning_rate": 0.0001872,
+      "loss": 2.4522,
       "step": 632
     },
     {
       "epoch": 1.0128,
-      "grad_norm": 5.944527626037598,
-      "learning_rate": 0.00025,
-      "loss": 2.0087,
+      "grad_norm": 6.208230018615723,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 2.3181,
       "step": 633
     },
     {
       "epoch": 1.0144,
-      "grad_norm": 5.785839557647705,
-      "learning_rate": 0.00025039999999999996,
-      "loss": 2.0019,
+      "grad_norm": 6.017099380493164,
+      "learning_rate": 0.00018779999999999998,
+      "loss": 2.4969,
       "step": 634
     },
     {
       "epoch": 1.016,
-      "grad_norm": 5.347036361694336,
-      "learning_rate": 0.00025079999999999997,
-      "loss": 2.0248,
+      "grad_norm": 5.968253135681152,
+      "learning_rate": 0.0001881,
+      "loss": 2.4739,
       "step": 635
     },
     {
       "epoch": 1.0176,
-      "grad_norm": 2.230567693710327,
-      "learning_rate": 0.0002512,
-      "loss": 2.3992,
+      "grad_norm": 5.338341236114502,
+      "learning_rate": 0.00018839999999999997,
+      "loss": 2.1392,
       "step": 636
     },
     {
       "epoch": 1.0192,
-      "grad_norm": 3.1952781677246094,
-      "learning_rate": 0.0002516,
-      "loss": 1.7205,
+      "grad_norm": 3.1507129669189453,
+      "learning_rate": 0.00018869999999999998,
+      "loss": 2.1685,
       "step": 637
     },
     {
       "epoch": 1.0208,
-      "grad_norm": 1.0791155099868774,
-      "learning_rate": 0.00025199999999999995,
-      "loss": 1.8243,
+      "grad_norm": 1.9297593832015991,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 1.8609,
       "step": 638
     },
     {
       "epoch": 1.0224,
-      "grad_norm": 1.0545350313186646,
-      "learning_rate": 0.0002524,
-      "loss": 1.6489,
+      "grad_norm": 1.8639192581176758,
+      "learning_rate": 0.0001893,
+      "loss": 2.1603,
       "step": 639
     },
     {
       "epoch": 1.024,
-      "grad_norm": 1.7690002918243408,
-      "learning_rate": 0.00025279999999999996,
-      "loss": 1.6945,
+      "grad_norm": 2.1058402061462402,
+      "learning_rate": 0.00018959999999999997,
+      "loss": 1.907,
       "step": 640
     },
     {
       "epoch": 1.0256,
-      "grad_norm": 1.6258310079574585,
-      "learning_rate": 0.0002532,
-      "loss": 1.7829,
+      "grad_norm": 1.7498681545257568,
+      "learning_rate": 0.00018989999999999998,
+      "loss": 1.9661,
       "step": 641
     },
     {
       "epoch": 1.0272,
-      "grad_norm": 1.9372260570526123,
-      "learning_rate": 0.0002536,
-      "loss": 1.6907,
+      "grad_norm": 2.575547218322754,
+      "learning_rate": 0.0001902,
+      "loss": 2.134,
       "step": 642
     },
     {
       "epoch": 1.0288,
-      "grad_norm": 1.9646409749984741,
-      "learning_rate": 0.000254,
-      "loss": 1.7792,
+      "grad_norm": 1.5888687372207642,
+      "learning_rate": 0.0001905,
+      "loss": 2.1148,
       "step": 643
     },
     {
       "epoch": 1.0304,
-      "grad_norm": 1.4375287294387817,
-      "learning_rate": 0.00025439999999999995,
-      "loss": 1.5501,
+      "grad_norm": 0.9597282409667969,
+      "learning_rate": 0.00019079999999999998,
+      "loss": 1.9374,
       "step": 644
     },
     {
       "epoch": 1.032,
-      "grad_norm": 2.31491756439209,
-      "learning_rate": 0.0002548,
-      "loss": 1.665,
+      "grad_norm": 1.114833116531372,
+      "learning_rate": 0.00019109999999999998,
+      "loss": 1.9728,
       "step": 645
     },
     {
       "epoch": 1.0336,
-      "grad_norm": 0.9905027747154236,
-      "learning_rate": 0.00025519999999999997,
-      "loss": 1.6621,
+      "grad_norm": 0.861680269241333,
+      "learning_rate": 0.0001914,
+      "loss": 1.9322,
       "step": 646
     },
     {
       "epoch": 1.0352,
-      "grad_norm": 1.7029163837432861,
-      "learning_rate": 0.0002556,
-      "loss": 1.852,
+      "grad_norm": 0.8410967588424683,
+      "learning_rate": 0.0001917,
+      "loss": 1.8089,
       "step": 647
     },
     {
       "epoch": 1.0368,
-      "grad_norm": 1.4749594926834106,
-      "learning_rate": 0.000256,
-      "loss": 1.8307,
+      "grad_norm": 0.9906688332557678,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 1.9472,
       "step": 648
     },
     {
       "epoch": 1.0384,
-      "grad_norm": 2.049144744873047,
-      "learning_rate": 0.0002564,
-      "loss": 1.8095,
+      "grad_norm": 0.9408037066459656,
+      "learning_rate": 0.00019229999999999999,
+      "loss": 1.9723,
       "step": 649
     },
     {
       "epoch": 1.04,
-      "grad_norm": 1.4582281112670898,
-      "learning_rate": 0.00025679999999999995,
-      "loss": 1.8413,
+      "grad_norm": 0.9392821788787842,
+      "learning_rate": 0.0001926,
+      "loss": 1.99,
       "step": 650
     },
     {
       "epoch": 1.0416,
-      "grad_norm": 1.0833181142807007,
-      "learning_rate": 0.00025719999999999996,
-      "loss": 1.7638,
+      "grad_norm": 1.3041807413101196,
+      "learning_rate": 0.0001929,
+      "loss": 1.9668,
       "step": 651
     },
     {
       "epoch": 1.0432,
-      "grad_norm": 1.0526366233825684,
-      "learning_rate": 0.0002576,
-      "loss": 1.8573,
+      "grad_norm": 1.1237937211990356,
+      "learning_rate": 0.00019319999999999998,
+      "loss": 1.9616,
       "step": 652
     },
     {
       "epoch": 1.0448,
-      "grad_norm": 0.9940699338912964,
-      "learning_rate": 0.000258,
-      "loss": 1.6459,
+      "grad_norm": 0.9516493082046509,
+      "learning_rate": 0.0001935,
+      "loss": 2.046,
       "step": 653
     },
     {
       "epoch": 1.0464,
-      "grad_norm": 0.8226042985916138,
-      "learning_rate": 0.00025839999999999994,
-      "loss": 1.6771,
+      "grad_norm": 0.9648451209068298,
+      "learning_rate": 0.0001938,
+      "loss": 2.0423,
       "step": 654
     },
     {
       "epoch": 1.048,
-      "grad_norm": 1.3881866931915283,
-      "learning_rate": 0.0002588,
-      "loss": 1.8241,
+      "grad_norm": 1.1745597124099731,
+      "learning_rate": 0.0001941,
+      "loss": 1.6941,
       "step": 655
     },
     {
       "epoch": 1.0496,
-      "grad_norm": 1.8918851613998413,
-      "learning_rate": 0.00025919999999999996,
-      "loss": 1.8669,
+      "grad_norm": 1.258222222328186,
+      "learning_rate": 0.00019439999999999998,
+      "loss": 1.8956,
       "step": 656
     },
     {
       "epoch": 1.0512,
-      "grad_norm": 1.265293836593628,
-      "learning_rate": 0.00025959999999999997,
-      "loss": 1.6453,
+      "grad_norm": 1.8220093250274658,
+      "learning_rate": 0.0001947,
+      "loss": 1.906,
       "step": 657
     },
     {
       "epoch": 1.0528,
-      "grad_norm": 1.5398906469345093,
-      "learning_rate": 0.00026,
-      "loss": 1.8982,
+      "grad_norm": 1.4318503141403198,
+      "learning_rate": 0.000195,
+      "loss": 1.9818,
       "step": 658
     },
     {
       "epoch": 1.0544,
-      "grad_norm": 1.0233063697814941,
-      "learning_rate": 0.0002604,
-      "loss": 1.6833,
+      "grad_norm": 1.7209746837615967,
+      "learning_rate": 0.00019529999999999998,
+      "loss": 1.7445,
       "step": 659
     },
     {
       "epoch": 1.056,
-      "grad_norm": 1.4029862880706787,
-      "learning_rate": 0.00026079999999999994,
-      "loss": 1.8216,
+      "grad_norm": 1.0792475938796997,
+      "learning_rate": 0.00019559999999999998,
+      "loss": 2.0065,
       "step": 660
     },
     {
       "epoch": 1.0576,
-      "grad_norm": 1.1457536220550537,
-      "learning_rate": 0.0002612,
-      "loss": 1.6201,
+      "grad_norm": 1.0406444072723389,
+      "learning_rate": 0.0001959,
+      "loss": 1.8579,
       "step": 661
     },
     {
       "epoch": 1.0592,
-      "grad_norm": 1.3890804052352905,
-      "learning_rate": 0.00026159999999999996,
-      "loss": 1.6713,
+      "grad_norm": 1.5529998540878296,
+      "learning_rate": 0.0001962,
+      "loss": 1.8712,
       "step": 662
     },
     {
       "epoch": 1.0608,
-      "grad_norm": 1.0882718563079834,
-      "learning_rate": 0.00026199999999999997,
-      "loss": 1.7714,
+      "grad_norm": 1.0818697214126587,
+      "learning_rate": 0.00019649999999999998,
+      "loss": 1.9727,
       "step": 663
     },
     {
       "epoch": 1.0624,
-      "grad_norm": 1.2843750715255737,
-      "learning_rate": 0.0002624,
-      "loss": 1.6399,
+      "grad_norm": 1.0074880123138428,
+      "learning_rate": 0.00019679999999999999,
+      "loss": 1.9209,
       "step": 664
     },
     {
       "epoch": 1.064,
-      "grad_norm": 1.7036292552947998,
-      "learning_rate": 0.0002628,
-      "loss": 2.1274,
+      "grad_norm": 2.4491970539093018,
+      "learning_rate": 0.0001971,
+      "loss": 2.1664,
       "step": 665
     },
     {
       "epoch": 1.0656,
-      "grad_norm": 2.123671293258667,
-      "learning_rate": 0.00026319999999999995,
-      "loss": 2.1099,
+      "grad_norm": 1.978613018989563,
+      "learning_rate": 0.0001974,
+      "loss": 2.3415,
       "step": 666
     },
     {
       "epoch": 1.0672,
-      "grad_norm": 1.3578087091445923,
-      "learning_rate": 0.0002636,
-      "loss": 1.6789,
+      "grad_norm": 1.2120779752731323,
+      "learning_rate": 0.00019769999999999998,
+      "loss": 2.3223,
       "step": 667
     },
     {
       "epoch": 1.0688,
-      "grad_norm": 2.183074951171875,
-      "learning_rate": 0.00026399999999999997,
-      "loss": 2.2433,
+      "grad_norm": 1.1476809978485107,
+      "learning_rate": 0.000198,
+      "loss": 2.0348,
       "step": 668
     },
     {
       "epoch": 1.0704,
-      "grad_norm": 1.4325931072235107,
-      "learning_rate": 0.0002644,
-      "loss": 2.0169,
+      "grad_norm": 1.0597435235977173,
+      "learning_rate": 0.0001983,
+      "loss": 2.0713,
       "step": 669
     },
     {
       "epoch": 1.072,
-      "grad_norm": 1.475628137588501,
-      "learning_rate": 0.0002648,
-      "loss": 2.0555,
+      "grad_norm": 1.3737848997116089,
+      "learning_rate": 0.0001986,
+      "loss": 2.0087,
       "step": 670
     },
     {
       "epoch": 1.0735999999999999,
-      "grad_norm": 1.7524343729019165,
-      "learning_rate": 0.0002652,
-      "loss": 1.9133,
+      "grad_norm": 1.1806519031524658,
+      "learning_rate": 0.00019889999999999998,
+      "loss": 1.9085,
       "step": 671
     },
     {
       "epoch": 1.0752,
-      "grad_norm": 1.3496627807617188,
-      "learning_rate": 0.00026559999999999995,
-      "loss": 1.8063,
+      "grad_norm": 1.1602704524993896,
+      "learning_rate": 0.0001992,
+      "loss": 2.383,
       "step": 672
     },
     {
       "epoch": 1.0768,
-      "grad_norm": 1.2976840734481812,
-      "learning_rate": 0.000266,
-      "loss": 2.256,
+      "grad_norm": 1.4072346687316895,
+      "learning_rate": 0.0001995,
+      "loss": 2.2611,
       "step": 673
     },
     {
       "epoch": 1.0784,
-      "grad_norm": 1.704746127128601,
-      "learning_rate": 0.00026639999999999997,
-      "loss": 1.7376,
+      "grad_norm": 1.908771276473999,
+      "learning_rate": 0.0001998,
+      "loss": 2.3743,
       "step": 674
     },
     {
       "epoch": 1.08,
-      "grad_norm": 2.0752077102661133,
-      "learning_rate": 0.0002668,
-      "loss": 2.7425,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001998,
+      "loss": 2.1281,
       "step": 675
     },
     {
       "epoch": 1.0816,
-      "grad_norm": 18.08973503112793,
-      "learning_rate": 0.0002672,
-      "loss": 3.7616,
+      "grad_norm": 19.099306106567383,
+      "learning_rate": 0.00020009999999999998,
+      "loss": 3.7561,
       "step": 676
     },
     {
       "epoch": 1.0832,
-      "grad_norm": 4.786888599395752,
-      "learning_rate": 0.0002676,
-      "loss": 2.572,
+      "grad_norm": 11.901508331298828,
+      "learning_rate": 0.0002004,
+      "loss": 3.0756,
       "step": 677
     },
     {
       "epoch": 1.0848,
-      "grad_norm": 4.977888584136963,
-      "learning_rate": 0.00026799999999999995,
-      "loss": 2.6192,
+      "grad_norm": 7.387051105499268,
+      "learning_rate": 0.0002007,
+      "loss": 2.6726,
       "step": 678
     },
     {
       "epoch": 1.0864,
-      "grad_norm": 1.7943665981292725,
-      "learning_rate": 0.0002684,
-      "loss": 2.0963,
+      "grad_norm": 2.313292980194092,
+      "learning_rate": 0.000201,
+      "loss": 2.0645,
       "step": 679
     },
     {
       "epoch": 1.088,
-      "grad_norm": 3.845595359802246,
-      "learning_rate": 0.0002688,
-      "loss": 2.101,
+      "grad_norm": 2.085894823074341,
+      "learning_rate": 0.0002013,
+      "loss": 2.043,
       "step": 680
     },
     {
       "epoch": 1.0896,
-      "grad_norm": 3.880871295928955,
-      "learning_rate": 0.0002692,
-      "loss": 2.0843,
+      "grad_norm": 3.863337278366089,
+      "learning_rate": 0.0002016,
+      "loss": 2.1174,
       "step": 681
     },
     {
       "epoch": 1.0912,
-      "grad_norm": 4.223239898681641,
-      "learning_rate": 0.00026959999999999994,
-      "loss": 2.2103,
+      "grad_norm": 4.193874835968018,
+      "learning_rate": 0.0002019,
+      "loss": 1.9003,
       "step": 682
     },
     {
       "epoch": 1.0928,
-      "grad_norm": 4.312764644622803,
-      "learning_rate": 0.00027,
-      "loss": 1.816,
+      "grad_norm": 4.875527858734131,
+      "learning_rate": 0.0002022,
+      "loss": 2.187,
       "step": 683
     },
     {
       "epoch": 1.0944,
-      "grad_norm": 2.8680918216705322,
-      "learning_rate": 0.00027039999999999996,
-      "loss": 1.5292,
+      "grad_norm": 2.392341375350952,
+      "learning_rate": 0.0002025,
+      "loss": 2.2106,
       "step": 684
     },
     {
       "epoch": 1.096,
-      "grad_norm": 5.548489093780518,
-      "learning_rate": 0.00027079999999999997,
-      "loss": 2.2242,
+      "grad_norm": 1.1724746227264404,
+      "learning_rate": 0.0002028,
+      "loss": 1.894,
       "step": 685
     },
     {
       "epoch": 1.0976,
-      "grad_norm": 0.9976663589477539,
-      "learning_rate": 0.0002712,
-      "loss": 1.8811,
+      "grad_norm": 1.3257207870483398,
+      "learning_rate": 0.0002031,
+      "loss": 1.8114,
       "step": 686
     },
     {
       "epoch": 1.0992,
-      "grad_norm": 1.6501846313476562,
-      "learning_rate": 0.0002716,
-      "loss": 1.5119,
+      "grad_norm": 1.1348336935043335,
+      "learning_rate": 0.00020339999999999998,
+      "loss": 1.6732,
       "step": 687
     },
     {
       "epoch": 1.1008,
-      "grad_norm": 0.8941024541854858,
-      "learning_rate": 0.00027199999999999994,
-      "loss": 1.4123,
+      "grad_norm": 2.1080403327941895,
+      "learning_rate": 0.0002037,
+      "loss": 1.9022,
       "step": 688
     },
     {
       "epoch": 1.1024,
-      "grad_norm": 1.797055721282959,
-      "learning_rate": 0.0002724,
-      "loss": 1.7278,
+      "grad_norm": 1.1208211183547974,
+      "learning_rate": 0.000204,
+      "loss": 1.6641,
       "step": 689
     },
     {
       "epoch": 1.104,
-      "grad_norm": 0.8527103066444397,
-      "learning_rate": 0.00027279999999999996,
-      "loss": 1.5845,
+      "grad_norm": 1.4055548906326294,
+      "learning_rate": 0.0002043,
+      "loss": 1.7557,
       "step": 690
     },
     {
       "epoch": 1.1056,
-      "grad_norm": 1.487897515296936,
-      "learning_rate": 0.00027319999999999997,
-      "loss": 1.5444,
+      "grad_norm": 1.0471781492233276,
+      "learning_rate": 0.00020459999999999999,
+      "loss": 1.629,
       "step": 691
     },
     {
       "epoch": 1.1072,
-      "grad_norm": 1.2457863092422485,
-      "learning_rate": 0.0002736,
-      "loss": 1.646,
+      "grad_norm": 0.8493616580963135,
+      "learning_rate": 0.0002049,
+      "loss": 1.8218,
       "step": 692
     },
     {
       "epoch": 1.1088,
-      "grad_norm": 1.3996880054473877,
-      "learning_rate": 0.000274,
-      "loss": 1.5098,
+      "grad_norm": 1.1957522630691528,
+      "learning_rate": 0.0002052,
+      "loss": 2.1286,
       "step": 693
     },
     {
       "epoch": 1.1104,
-      "grad_norm": 1.7704321146011353,
-      "learning_rate": 0.00027439999999999995,
-      "loss": 1.6845,
+      "grad_norm": 0.6791967153549194,
+      "learning_rate": 0.0002055,
+      "loss": 1.8816,
       "step": 694
     },
     {
       "epoch": 1.112,
-      "grad_norm": 1.5547606945037842,
-      "learning_rate": 0.0002748,
-      "loss": 1.7829,
+      "grad_norm": 0.8928971290588379,
+      "learning_rate": 0.0002058,
+      "loss": 1.8552,
       "step": 695
     },
     {
       "epoch": 1.1136,
-      "grad_norm": 1.3222217559814453,
-      "learning_rate": 0.00027519999999999997,
-      "loss": 1.5888,
+      "grad_norm": 1.1113934516906738,
+      "learning_rate": 0.0002061,
+      "loss": 1.8408,
       "step": 696
     },
     {
       "epoch": 1.1152,
-      "grad_norm": 0.9179243445396423,
-      "learning_rate": 0.0002756,
-      "loss": 1.7316,
+      "grad_norm": 0.948301374912262,
+      "learning_rate": 0.00020639999999999998,
+      "loss": 1.918,
       "step": 697
     },
     {
       "epoch": 1.1168,
-      "grad_norm": 1.142059087753296,
-      "learning_rate": 0.000276,
-      "loss": 1.7593,
+      "grad_norm": 0.7305535078048706,
+      "learning_rate": 0.00020669999999999996,
+      "loss": 1.722,
       "step": 698
     },
     {
       "epoch": 1.1184,
-      "grad_norm": 1.3432698249816895,
-      "learning_rate": 0.0002764,
-      "loss": 1.7522,
+      "grad_norm": 2.527282953262329,
+      "learning_rate": 0.00020699999999999996,
+      "loss": 2.1528,
       "step": 699
     },
     {
       "epoch": 1.12,
-      "grad_norm": 0.8155900239944458,
-      "learning_rate": 0.00027679999999999995,
-      "loss": 1.7013,
-      "step": 700
-    },
-    {
-      "epoch": 1.12,
-      "eval_cer": 0.44886196037788495,
-      "eval_loss": 2.1150970458984375,
-      "eval_runtime": 158.4276,
-      "eval_samples_per_second": 19.795,
-      "eval_steps_per_second": 1.237,
-      "eval_wer": 0.6820018650917004,
+      "grad_norm": 2.323239326477051,
+      "learning_rate": 0.00020729999999999997,
+      "loss": 1.8547,
       "step": 700
     },
     {
       "epoch": 1.1216,
-      "grad_norm": 0.9745352864265442,
-      "learning_rate": 0.0002772,
-      "loss": 1.7217,
+      "grad_norm": 1.690564751625061,
+      "learning_rate": 0.00020759999999999998,
+      "loss": 1.7551,
       "step": 701
     },
     {
       "epoch": 1.1232,
-      "grad_norm": 0.939247190952301,
-      "learning_rate": 0.00027759999999999997,
-      "loss": 1.5636,
+      "grad_norm": 1.1845993995666504,
+      "learning_rate": 0.00020789999999999996,
+      "loss": 2.0184,
       "step": 702
     },
     {
       "epoch": 1.1248,
-      "grad_norm": 1.1377625465393066,
-      "learning_rate": 0.000278,
-      "loss": 1.3686,
+      "grad_norm": 0.9866268038749695,
+      "learning_rate": 0.00020819999999999996,
+      "loss": 1.7094,
       "step": 703
     },
     {
       "epoch": 1.1264,
-      "grad_norm": 1.3641624450683594,
-      "learning_rate": 0.0002784,
-      "loss": 1.6273,
+      "grad_norm": 1.6895757913589478,
+      "learning_rate": 0.00020849999999999997,
+      "loss": 1.805,
       "step": 704
     },
     {
       "epoch": 1.1280000000000001,
-      "grad_norm": 1.7075438499450684,
-      "learning_rate": 0.0002788,
-      "loss": 1.6859,
+      "grad_norm": 1.2320947647094727,
+      "learning_rate": 0.00020879999999999998,
+      "loss": 1.6979,
       "step": 705
     },
     {
       "epoch": 1.1296,
-      "grad_norm": 0.9067268371582031,
-      "learning_rate": 0.00027919999999999996,
-      "loss": 1.8782,
+      "grad_norm": 1.0378570556640625,
+      "learning_rate": 0.00020909999999999996,
+      "loss": 1.6646,
       "step": 706
     },
     {
       "epoch": 1.1312,
-      "grad_norm": 2.020303964614868,
-      "learning_rate": 0.00027959999999999997,
-      "loss": 1.8752,
+      "grad_norm": 1.7210016250610352,
+      "learning_rate": 0.00020939999999999997,
+      "loss": 2.0604,
       "step": 707
     },
     {
       "epoch": 1.1328,
-      "grad_norm": 1.2338893413543701,
-      "learning_rate": 0.00028,
-      "loss": 1.4884,
+      "grad_norm": 1.0748008489608765,
+      "learning_rate": 0.00020969999999999997,
+      "loss": 1.7569,
       "step": 708
     },
     {
       "epoch": 1.1344,
-      "grad_norm": 1.5765209197998047,
-      "learning_rate": 0.0002804,
-      "loss": 1.8257,
+      "grad_norm": 1.0989258289337158,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 1.6594,
       "step": 709
     },
     {
       "epoch": 1.1360000000000001,
-      "grad_norm": 1.7421718835830688,
-      "learning_rate": 0.0002808,
-      "loss": 1.993,
+      "grad_norm": 1.1351910829544067,
+      "learning_rate": 0.00021029999999999996,
+      "loss": 1.9019,
       "step": 710
     },
     {
       "epoch": 1.1376,
-      "grad_norm": 1.4602911472320557,
-      "learning_rate": 0.0002812,
-      "loss": 1.2757,
+      "grad_norm": 1.115064263343811,
+      "learning_rate": 0.00021059999999999997,
+      "loss": 1.6088,
       "step": 711
     },
     {
       "epoch": 1.1392,
-      "grad_norm": 1.140167236328125,
-      "learning_rate": 0.00028159999999999996,
-      "loss": 1.7157,
+      "grad_norm": 1.037343978881836,
+      "learning_rate": 0.00021089999999999998,
+      "loss": 1.5529,
       "step": 712
     },
     {
       "epoch": 1.1408,
-      "grad_norm": 3.379197120666504,
-      "learning_rate": 0.00028199999999999997,
-      "loss": 1.906,
+      "grad_norm": 1.8999748229980469,
+      "learning_rate": 0.00021119999999999996,
+      "loss": 1.7772,
       "step": 713
     },
     {
       "epoch": 1.1424,
-      "grad_norm": 1.0864256620407104,
-      "learning_rate": 0.0002824,
-      "loss": 1.6679,
+      "grad_norm": 1.6710184812545776,
+      "learning_rate": 0.00021149999999999996,
+      "loss": 1.7199,
       "step": 714
     },
     {
       "epoch": 1.144,
-      "grad_norm": 4.162722587585449,
-      "learning_rate": 0.0002828,
-      "loss": 1.85,
+      "grad_norm": 0.9634356498718262,
+      "learning_rate": 0.00021179999999999997,
+      "loss": 1.9266,
       "step": 715
     },
     {
       "epoch": 1.1456,
-      "grad_norm": 1.3800784349441528,
-      "learning_rate": 0.00028319999999999994,
-      "loss": 1.596,
+      "grad_norm": 1.2477611303329468,
+      "learning_rate": 0.00021209999999999998,
+      "loss": 1.7878,
       "step": 716
     },
     {
       "epoch": 1.1472,
-      "grad_norm": 1.3533785343170166,
-      "learning_rate": 0.0002836,
-      "loss": 1.3124,
+      "grad_norm": 1.0142041444778442,
+      "learning_rate": 0.00021239999999999996,
+      "loss": 1.4408,
       "step": 717
     },
     {
       "epoch": 1.1488,
-      "grad_norm": 1.9931690692901611,
-      "learning_rate": 0.00028399999999999996,
-      "loss": 2.2037,
+      "grad_norm": 4.752976417541504,
+      "learning_rate": 0.00021269999999999997,
+      "loss": 2.5672,
       "step": 718
     },
     {
       "epoch": 1.1504,
-      "grad_norm": 1.353873372077942,
-      "learning_rate": 0.0002844,
-      "loss": 1.7149,
+      "grad_norm": 1.2975269556045532,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 1.8788,
       "step": 719
     },
     {
       "epoch": 1.152,
-      "grad_norm": 1.64341139793396,
-      "learning_rate": 0.0002848,
-      "loss": 1.9453,
+      "grad_norm": 1.2199846506118774,
+      "learning_rate": 0.00021329999999999998,
+      "loss": 1.8968,
       "step": 720
     },
     {
       "epoch": 1.1536,
-      "grad_norm": 1.1980323791503906,
-      "learning_rate": 0.0002852,
-      "loss": 1.8312,
+      "grad_norm": 1.3590461015701294,
+      "learning_rate": 0.00021359999999999996,
+      "loss": 1.5399,
       "step": 721
     },
     {
       "epoch": 1.1552,
-      "grad_norm": 2.0850045680999756,
-      "learning_rate": 0.00028559999999999995,
-      "loss": 1.5216,
+      "grad_norm": 3.0514047145843506,
+      "learning_rate": 0.00021389999999999997,
+      "loss": 2.1824,
       "step": 722
     },
     {
       "epoch": 1.1568,
-      "grad_norm": 1.8492074012756348,
-      "learning_rate": 0.00028599999999999996,
-      "loss": 2.448,
+      "grad_norm": 3.324256658554077,
+      "learning_rate": 0.00021419999999999998,
+      "loss": 2.248,
       "step": 723
     },
     {
       "epoch": 1.1584,
-      "grad_norm": 1.954471468925476,
-      "learning_rate": 0.00028639999999999997,
-      "loss": 1.6491,
+      "grad_norm": 1.6645241975784302,
+      "learning_rate": 0.00021449999999999998,
+      "loss": 2.0545,
       "step": 724
     },
     {
       "epoch": 1.16,
-      "grad_norm": 1.8707425594329834,
-      "learning_rate": 0.0002868,
-      "loss": 1.9693,
+      "grad_norm": 3.6503653526306152,
+      "learning_rate": 0.00021479999999999996,
+      "loss": 2.4739,
       "step": 725
     },
     {
       "epoch": 1.1616,
-      "grad_norm": 28.85036277770996,
-      "learning_rate": 0.0002872,
-      "loss": 4.7933,
+      "grad_norm": 14.17703914642334,
+      "learning_rate": 0.00021509999999999997,
+      "loss": 3.3052,
       "step": 726
     },
     {
       "epoch": 1.1632,
-      "grad_norm": 4.971120834350586,
-      "learning_rate": 0.0002876,
-      "loss": 2.2725,
+      "grad_norm": 12.942245483398438,
+      "learning_rate": 0.00021539999999999998,
+      "loss": 3.0785,
       "step": 727
     },
     {
       "epoch": 1.1648,
-      "grad_norm": 11.561688423156738,
-      "learning_rate": 0.00028799999999999995,
-      "loss": 3.1344,
+      "grad_norm": 3.433819055557251,
+      "learning_rate": 0.00021569999999999998,
+      "loss": 1.9356,
       "step": 728
     },
     {
       "epoch": 1.1663999999999999,
-      "grad_norm": 2.439958333969116,
-      "learning_rate": 0.00028839999999999996,
-      "loss": 1.7422,
+      "grad_norm": 3.700389862060547,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 2.3045,
       "step": 729
     },
     {
       "epoch": 1.168,
-      "grad_norm": 2.66070818901062,
-      "learning_rate": 0.00028879999999999997,
-      "loss": 2.2646,
+      "grad_norm": 4.333859443664551,
+      "learning_rate": 0.00021629999999999997,
+      "loss": 2.1794,
       "step": 730
     },
     {
       "epoch": 1.1696,
-      "grad_norm": 1.5238914489746094,
-      "learning_rate": 0.0002892,
-      "loss": 1.971,
+      "grad_norm": 1.7107884883880615,
+      "learning_rate": 0.00021659999999999998,
+      "loss": 2.3069,
       "step": 731
     },
     {
       "epoch": 1.1712,
-      "grad_norm": 5.540120601654053,
-      "learning_rate": 0.0002896,
-      "loss": 1.7024,
+      "grad_norm": 1.8687323331832886,
+      "learning_rate": 0.0002169,
+      "loss": 2.7628,
       "step": 732
     },
     {
       "epoch": 1.1728,
-      "grad_norm": 5.015547275543213,
-      "learning_rate": 0.00029,
-      "loss": 1.9131,
+      "grad_norm": 4.187427043914795,
+      "learning_rate": 0.00021719999999999997,
+      "loss": 1.8492,
       "step": 733
     },
     {
       "epoch": 1.1743999999999999,
-      "grad_norm": 7.225318431854248,
-      "learning_rate": 0.00029039999999999996,
-      "loss": 2.0965,
+      "grad_norm": 6.42722225189209,
+      "learning_rate": 0.00021749999999999997,
+      "loss": 1.9605,
       "step": 734
     },
     {
       "epoch": 1.176,
-      "grad_norm": 5.978356838226318,
-      "learning_rate": 0.00029079999999999997,
-      "loss": 1.9004,
+      "grad_norm": 5.926379203796387,
+      "learning_rate": 0.00021779999999999998,
+      "loss": 1.9447,
       "step": 735
     },
     {
       "epoch": 1.1776,
-      "grad_norm": 5.611774444580078,
-      "learning_rate": 0.0002912,
-      "loss": 2.0044,
+      "grad_norm": 5.0843119621276855,
+      "learning_rate": 0.00021809999999999996,
+      "loss": 1.9547,
       "step": 736
     },
     {
       "epoch": 1.1792,
-      "grad_norm": 3.98929500579834,
-      "learning_rate": 0.0002916,
-      "loss": 1.827,
+      "grad_norm": 3.0548288822174072,
+      "learning_rate": 0.00021839999999999997,
+      "loss": 1.9428,
       "step": 737
     },
     {
       "epoch": 1.1808,
-      "grad_norm": 1.499614953994751,
-      "learning_rate": 0.000292,
-      "loss": 1.7716,
+      "grad_norm": 1.9500654935836792,
+      "learning_rate": 0.00021869999999999998,
+      "loss": 1.707,
       "step": 738
     },
     {
       "epoch": 1.1824,
-      "grad_norm": 0.908250093460083,
-      "learning_rate": 0.0002924,
-      "loss": 1.751,
+      "grad_norm": 1.0856446027755737,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 1.8138,
       "step": 739
     },
     {
       "epoch": 1.184,
-      "grad_norm": 1.4749705791473389,
-      "learning_rate": 0.00029279999999999996,
-      "loss": 1.7102,
+      "grad_norm": 2.099169969558716,
+      "learning_rate": 0.00021929999999999996,
+      "loss": 1.9153,
       "step": 740
     },
     {
       "epoch": 1.1856,
-      "grad_norm": 2.4379866123199463,
-      "learning_rate": 0.00029319999999999997,
-      "loss": 1.5624,
+      "grad_norm": 3.320614814758301,
+      "learning_rate": 0.00021959999999999997,
+      "loss": 1.9022,
       "step": 741
     },
     {
       "epoch": 1.1872,
-      "grad_norm": 1.296765923500061,
-      "learning_rate": 0.0002936,
-      "loss": 1.6655,
+      "grad_norm": 2.1835997104644775,
+      "learning_rate": 0.00021989999999999998,
+      "loss": 1.7657,
       "step": 742
     },
     {
       "epoch": 1.1888,
-      "grad_norm": 1.7663196325302124,
-      "learning_rate": 0.000294,
-      "loss": 1.4732,
+      "grad_norm": 1.8994768857955933,
+      "learning_rate": 0.00022019999999999999,
+      "loss": 1.6526,
       "step": 743
     },
     {
       "epoch": 1.1904,
-      "grad_norm": 1.1070098876953125,
-      "learning_rate": 0.00029439999999999995,
-      "loss": 1.6552,
+      "grad_norm": 1.6875131130218506,
+      "learning_rate": 0.00022049999999999997,
+      "loss": 1.7326,
       "step": 744
     },
     {
       "epoch": 1.192,
-      "grad_norm": 0.7955397367477417,
-      "learning_rate": 0.00029479999999999996,
-      "loss": 1.335,
+      "grad_norm": 1.7935808897018433,
+      "learning_rate": 0.00022079999999999997,
+      "loss": 1.6772,
       "step": 745
     },
     {
       "epoch": 1.1936,
-      "grad_norm": 1.6219990253448486,
-      "learning_rate": 0.00029519999999999997,
-      "loss": 1.5489,
+      "grad_norm": 1.137858510017395,
+      "learning_rate": 0.00022109999999999998,
+      "loss": 1.8035,
       "step": 746
     },
     {
       "epoch": 1.1952,
-      "grad_norm": 0.8686781525611877,
-      "learning_rate": 0.0002956,
-      "loss": 1.417,
+      "grad_norm": 1.6366093158721924,
+      "learning_rate": 0.0002214,
+      "loss": 1.6763,
       "step": 747
     },
     {
       "epoch": 1.1968,
-      "grad_norm": 2.419113874435425,
-      "learning_rate": 0.000296,
-      "loss": 1.6413,
+      "grad_norm": 1.0513057708740234,
+      "learning_rate": 0.00022169999999999997,
+      "loss": 2.0465,
       "step": 748
     },
     {
       "epoch": 1.1984,
-      "grad_norm": 2.114264488220215,
-      "learning_rate": 0.0002964,
-      "loss": 1.9603,
+      "grad_norm": 2.0557422637939453,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 1.6699,
       "step": 749
     },
     {
       "epoch": 1.2,
-      "grad_norm": 1.2210099697113037,
-      "learning_rate": 0.00029679999999999995,
-      "loss": 1.8247,
+      "grad_norm": 2.343322515487671,
+      "learning_rate": 0.00022229999999999998,
+      "loss": 1.5154,
       "step": 750
     },
     {
       "epoch": 1.2016,
-      "grad_norm": 1.4565551280975342,
-      "learning_rate": 0.00029719999999999996,
-      "loss": 1.3969,
+      "grad_norm": 2.144352912902832,
+      "learning_rate": 0.0002226,
+      "loss": 1.9019,
       "step": 751
     },
     {
       "epoch": 1.2032,
-      "grad_norm": 0.8622385263442993,
-      "learning_rate": 0.00029759999999999997,
-      "loss": 1.4452,
+      "grad_norm": 1.0311518907546997,
+      "learning_rate": 0.00022289999999999997,
+      "loss": 1.611,
       "step": 752
     },
     {
       "epoch": 1.2048,
-      "grad_norm": 0.817267656326294,
-      "learning_rate": 0.000298,
-      "loss": 1.8705,
+      "grad_norm": 1.594190001487732,
+      "learning_rate": 0.00022319999999999998,
+      "loss": 1.8844,
       "step": 753
     },
     {
       "epoch": 1.2064,
-      "grad_norm": 1.0307005643844604,
-      "learning_rate": 0.0002984,
-      "loss": 1.4181,
+      "grad_norm": 1.0888090133666992,
+      "learning_rate": 0.00022349999999999998,
+      "loss": 1.4243,
       "step": 754
     },
     {
       "epoch": 1.208,
-      "grad_norm": 1.6527661085128784,
-      "learning_rate": 0.0002988,
-      "loss": 1.6909,
+      "grad_norm": 0.8538291454315186,
+      "learning_rate": 0.0002238,
+      "loss": 1.7549,
       "step": 755
     },
     {
       "epoch": 1.2096,
-      "grad_norm": 1.1635178327560425,
-      "learning_rate": 0.00029919999999999995,
-      "loss": 1.6779,
+      "grad_norm": 0.9313039779663086,
+      "learning_rate": 0.00022409999999999997,
+      "loss": 1.5294,
       "step": 756
     },
     {
       "epoch": 1.2112,
-      "grad_norm": 1.4051568508148193,
-      "learning_rate": 0.00029959999999999996,
-      "loss": 1.2773,
+      "grad_norm": 1.0965137481689453,
+      "learning_rate": 0.00022439999999999998,
+      "loss": 1.5921,
       "step": 757
     },
     {
       "epoch": 1.2128,
-      "grad_norm": 1.8025474548339844,
-      "learning_rate": 0.0003,
-      "loss": 1.7951,
+      "grad_norm": 1.8259750604629517,
+      "learning_rate": 0.0002247,
+      "loss": 1.6646,
       "step": 758
     },
     {
       "epoch": 1.2144,
-      "grad_norm": 1.042132019996643,
-      "learning_rate": 0.0003004,
-      "loss": 1.5419,
+      "grad_norm": 1.3318666219711304,
+      "learning_rate": 0.000225,
+      "loss": 1.7923,
       "step": 759
     },
     {
       "epoch": 1.216,
-      "grad_norm": 1.1578173637390137,
-      "learning_rate": 0.00030079999999999994,
-      "loss": 1.4649,
+      "grad_norm": 1.7984483242034912,
+      "learning_rate": 0.00022529999999999997,
+      "loss": 2.1313,
       "step": 760
     },
     {
       "epoch": 1.2176,
-      "grad_norm": 2.2228667736053467,
-      "learning_rate": 0.00030119999999999995,
-      "loss": 2.0639,
+      "grad_norm": 1.0261139869689941,
+      "learning_rate": 0.00022559999999999998,
+      "loss": 1.8728,
       "step": 761
     },
     {
       "epoch": 1.2192,
-      "grad_norm": 0.8677162528038025,
-      "learning_rate": 0.0003016,
-      "loss": 1.4764,
+      "grad_norm": 1.089312195777893,
+      "learning_rate": 0.0002259,
+      "loss": 1.7172,
       "step": 762
     },
     {
       "epoch": 1.2208,
-      "grad_norm": 1.2380160093307495,
-      "learning_rate": 0.00030199999999999997,
-      "loss": 1.6447,
+      "grad_norm": 0.9745715856552124,
+      "learning_rate": 0.00022619999999999997,
+      "loss": 1.8426,
       "step": 763
     },
     {
       "epoch": 1.2224,
-      "grad_norm": 1.3636889457702637,
-      "learning_rate": 0.0003024,
-      "loss": 1.7581,
+      "grad_norm": 1.1948559284210205,
+      "learning_rate": 0.00022649999999999998,
+      "loss": 1.6767,
       "step": 764
     },
     {
       "epoch": 1.224,
-      "grad_norm": 1.1308799982070923,
-      "learning_rate": 0.0003028,
-      "loss": 1.7186,
+      "grad_norm": 1.2432363033294678,
+      "learning_rate": 0.00022679999999999998,
+      "loss": 1.6967,
       "step": 765
     },
     {
       "epoch": 1.2256,
-      "grad_norm": 1.5796843767166138,
-      "learning_rate": 0.00030319999999999994,
-      "loss": 1.7703,
+      "grad_norm": 1.209937572479248,
+      "learning_rate": 0.0002271,
+      "loss": 2.2466,
       "step": 766
     },
     {
       "epoch": 1.2272,
-      "grad_norm": 1.0818617343902588,
-      "learning_rate": 0.00030359999999999995,
-      "loss": 1.418,
+      "grad_norm": 1.1278433799743652,
+      "learning_rate": 0.00022739999999999997,
+      "loss": 1.8898,
       "step": 767
     },
     {
       "epoch": 1.2288000000000001,
-      "grad_norm": 2.3142552375793457,
-      "learning_rate": 0.000304,
-      "loss": 1.935,
+      "grad_norm": 0.9980478286743164,
+      "learning_rate": 0.00022769999999999998,
+      "loss": 1.8697,
       "step": 768
     },
     {
       "epoch": 1.2304,
-      "grad_norm": 1.6650938987731934,
-      "learning_rate": 0.00030439999999999997,
-      "loss": 2.1436,
+      "grad_norm": 1.7807966470718384,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 1.6999,
       "step": 769
     },
     {
       "epoch": 1.232,
-      "grad_norm": 1.7241698503494263,
-      "learning_rate": 0.0003048,
-      "loss": 1.6536,
+      "grad_norm": 1.3307067155838013,
+      "learning_rate": 0.0002283,
+      "loss": 2.0235,
       "step": 770
     },
     {
       "epoch": 1.2336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0003048,
-      "loss": 1.5988,
+      "grad_norm": 1.336665391921997,
+      "learning_rate": 0.00022859999999999997,
+      "loss": 1.9368,
       "step": 771
     },
     {
       "epoch": 1.2352,
-      "grad_norm": 1.0828837156295776,
-      "learning_rate": 0.0003052,
-      "loss": 2.0584,
+      "grad_norm": 1.5342364311218262,
+      "learning_rate": 0.00022889999999999998,
+      "loss": 1.736,
       "step": 772
     },
     {
       "epoch": 1.2368000000000001,
-      "grad_norm": 1.8787623643875122,
-      "learning_rate": 0.00030559999999999995,
-      "loss": 2.1669,
+      "grad_norm": 1.804161548614502,
+      "learning_rate": 0.0002292,
+      "loss": 2.3303,
       "step": 773
     },
     {
       "epoch": 1.2384,
-      "grad_norm": 1.8352808952331543,
-      "learning_rate": 0.00030599999999999996,
-      "loss": 2.4238,
+      "grad_norm": 1.1865055561065674,
+      "learning_rate": 0.0002295,
+      "loss": 1.8258,
       "step": 774
     },
     {
       "epoch": 1.24,
-      "grad_norm": 2.0836939811706543,
-      "learning_rate": 0.0003064,
-      "loss": 2.7034,
+      "grad_norm": NaN,
+      "learning_rate": 0.0002295,
+      "loss": 1.4765,
       "step": 775
     },
     {
       "epoch": 1.2416,
-      "grad_norm": 7.965085983276367,
-      "learning_rate": 0.0003068,
-      "loss": 2.6308,
+      "grad_norm": 11.694743156433105,
+      "learning_rate": 0.00022979999999999997,
+      "loss": 3.1048,
       "step": 776
     },
     {
       "epoch": 1.2432,
-      "grad_norm": 2.8979556560516357,
-      "learning_rate": 0.0003072,
-      "loss": 1.8492,
+      "grad_norm": 4.708855628967285,
+      "learning_rate": 0.00023009999999999998,
+      "loss": 2.2849,
       "step": 777
     },
     {
       "epoch": 1.2448,
-      "grad_norm": 1.722564697265625,
-      "learning_rate": 0.0003076,
-      "loss": 1.5315,
+      "grad_norm": 1.5170272588729858,
+      "learning_rate": 0.0002304,
+      "loss": 2.1169,
       "step": 778
     },
     {
       "epoch": 1.2464,
-      "grad_norm": 2.6675570011138916,
-      "learning_rate": 0.00030799999999999995,
-      "loss": 2.2452,
+      "grad_norm": 3.2617526054382324,
+      "learning_rate": 0.0002307,
+      "loss": 2.6471,
       "step": 779
     },
     {
       "epoch": 1.248,
-      "grad_norm": 2.365095853805542,
-      "learning_rate": 0.00030839999999999996,
-      "loss": 1.8645,
+      "grad_norm": 2.189706325531006,
+      "learning_rate": 0.00023099999999999998,
+      "loss": 2.264,
       "step": 780
     },
     {
       "epoch": 1.2496,
-      "grad_norm": 2.9345123767852783,
-      "learning_rate": 0.0003088,
-      "loss": 1.9889,
+      "grad_norm": 2.7677061557769775,
+      "learning_rate": 0.00023129999999999998,
+      "loss": 2.2306,
       "step": 781
     },
     {
       "epoch": 1.2511999999999999,
-      "grad_norm": 2.7510879039764404,
-      "learning_rate": 0.0003092,
-      "loss": 1.863,
+      "grad_norm": 5.468600749969482,
+      "learning_rate": 0.0002316,
+      "loss": 2.0234,
       "step": 782
     },
     {
       "epoch": 1.2528000000000001,
-      "grad_norm": 3.8311636447906494,
-      "learning_rate": 0.0003096,
-      "loss": 1.7916,
+      "grad_norm": 6.928030014038086,
+      "learning_rate": 0.0002319,
+      "loss": 1.947,
       "step": 783
     },
     {
       "epoch": 1.2544,
-      "grad_norm": 2.39135479927063,
-      "learning_rate": 0.00031,
-      "loss": 1.5686,
+      "grad_norm": 7.065882205963135,
+      "learning_rate": 0.00023219999999999998,
+      "loss": 1.9462,
       "step": 784
     },
     {
       "epoch": 1.256,
-      "grad_norm": 2.308666944503784,
-      "learning_rate": 0.00031039999999999996,
-      "loss": 1.7752,
+      "grad_norm": 5.484043121337891,
+      "learning_rate": 0.00023249999999999999,
+      "loss": 2.0667,
       "step": 785
     },
     {
       "epoch": 1.2576,
-      "grad_norm": 1.1244654655456543,
-      "learning_rate": 0.00031079999999999997,
-      "loss": 1.5987,
+      "grad_norm": 4.054402828216553,
+      "learning_rate": 0.0002328,
+      "loss": 1.8082,
       "step": 786
     },
     {
       "epoch": 1.2591999999999999,
-      "grad_norm": 2.477769613265991,
-      "learning_rate": 0.00031120000000000003,
-      "loss": 1.5174,
+      "grad_norm": 2.286044120788574,
+      "learning_rate": 0.00023309999999999997,
+      "loss": 1.807,
       "step": 787
     },
     {
       "epoch": 1.2608,
-      "grad_norm": 1.5737215280532837,
-      "learning_rate": 0.0003116,
-      "loss": 1.4032,
+      "grad_norm": 1.4100284576416016,
+      "learning_rate": 0.00023339999999999998,
+      "loss": 1.7669,
       "step": 788
     },
     {
       "epoch": 1.2624,
-      "grad_norm": 1.4628270864486694,
-      "learning_rate": 0.000312,
-      "loss": 1.3916,
+      "grad_norm": 2.072185754776001,
+      "learning_rate": 0.0002337,
+      "loss": 1.549,
       "step": 789
     },
     {
       "epoch": 1.264,
-      "grad_norm": 2.7115325927734375,
-      "learning_rate": 0.0003124,
-      "loss": 1.2787,
+      "grad_norm": 2.230264663696289,
+      "learning_rate": 0.000234,
+      "loss": 1.6862,
       "step": 790
     },
     {
       "epoch": 1.2656,
-      "grad_norm": 2.164787769317627,
-      "learning_rate": 0.00031279999999999996,
-      "loss": 1.4398,
+      "grad_norm": 3.1086905002593994,
+      "learning_rate": 0.00023429999999999998,
+      "loss": 1.593,
       "step": 791
     },
     {
       "epoch": 1.2671999999999999,
-      "grad_norm": 1.2060961723327637,
-      "learning_rate": 0.00031319999999999997,
-      "loss": 1.4638,
+      "grad_norm": 2.1213574409484863,
+      "learning_rate": 0.00023459999999999998,
+      "loss": 1.5835,
       "step": 792
     },
     {
       "epoch": 1.2688,
-      "grad_norm": 1.0704708099365234,
-      "learning_rate": 0.0003135999999999999,
-      "loss": 1.4506,
+      "grad_norm": 4.03560209274292,
+      "learning_rate": 0.0002349,
+      "loss": 2.1431,
       "step": 793
     },
     {
       "epoch": 1.2704,
-      "grad_norm": 1.1875430345535278,
-      "learning_rate": 0.000314,
-      "loss": 1.6161,
+      "grad_norm": 1.5937985181808472,
+      "learning_rate": 0.0002352,
+      "loss": 1.8309,
       "step": 794
     },
     {
       "epoch": 1.272,
-      "grad_norm": 0.8472498655319214,
-      "learning_rate": 0.0003144,
-      "loss": 1.4842,
+      "grad_norm": 1.3385921716690063,
+      "learning_rate": 0.00023549999999999998,
+      "loss": 1.6525,
       "step": 795
     },
     {
       "epoch": 1.2736,
-      "grad_norm": 1.229743242263794,
-      "learning_rate": 0.00031479999999999995,
-      "loss": 1.5075,
+      "grad_norm": 1.2098594903945923,
+      "learning_rate": 0.00023579999999999999,
+      "loss": 1.849,
       "step": 796
     },
     {
       "epoch": 1.2752,
-      "grad_norm": 0.8419369459152222,
-      "learning_rate": 0.00031519999999999996,
-      "loss": 1.3735,
+      "grad_norm": 1.2983063459396362,
+      "learning_rate": 0.0002361,
+      "loss": 1.6045,
       "step": 797
     },
     {
       "epoch": 1.2768,
-      "grad_norm": 1.6676976680755615,
-      "learning_rate": 0.0003156,
-      "loss": 1.5081,
+      "grad_norm": 1.0883206129074097,
+      "learning_rate": 0.0002364,
+      "loss": 1.5005,
       "step": 798
     },
     {
       "epoch": 1.2784,
-      "grad_norm": 1.597878098487854,
-      "learning_rate": 0.00031599999999999993,
-      "loss": 1.513,
+      "grad_norm": 1.3664700984954834,
+      "learning_rate": 0.00023669999999999998,
+      "loss": 1.5107,
       "step": 799
     },
     {
       "epoch": 1.28,
-      "grad_norm": 0.893544614315033,
-      "learning_rate": 0.00031639999999999994,
-      "loss": 1.2995,
-      "step": 800
-    },
-    {
-      "epoch": 1.28,
-      "eval_cer": 0.415968429864073,
-      "eval_loss": 2.0757076740264893,
-      "eval_runtime": 158.58,
-      "eval_samples_per_second": 19.776,
-      "eval_steps_per_second": 1.236,
-      "eval_wer": 0.6244948709978241,
+      "grad_norm": 1.9094866514205933,
+      "learning_rate": 0.000237,
+      "loss": 1.8143,
       "step": 800
     },
     {
       "epoch": 1.2816,
-      "grad_norm": 3.154768466949463,
-      "learning_rate": 0.0003168,
-      "loss": 1.7691,
+      "grad_norm": 1.615760326385498,
+      "learning_rate": 0.0002373,
+      "loss": 1.889,
       "step": 801
     },
     {
       "epoch": 1.2832,
-      "grad_norm": 0.7817122340202332,
-      "learning_rate": 0.00031719999999999996,
-      "loss": 1.5512,
+      "grad_norm": 1.3856728076934814,
+      "learning_rate": 0.0002376,
+      "loss": 1.7875,
       "step": 802
     },
     {
       "epoch": 1.2848,
-      "grad_norm": 1.3338487148284912,
-      "learning_rate": 0.00031759999999999997,
-      "loss": 1.6868,
+      "grad_norm": 1.2767575979232788,
+      "learning_rate": 0.00023789999999999998,
+      "loss": 1.7884,
       "step": 803
     },
     {
       "epoch": 1.2864,
-      "grad_norm": 1.1260954141616821,
-      "learning_rate": 0.000318,
-      "loss": 1.6486,
+      "grad_norm": 0.93840491771698,
+      "learning_rate": 0.0002382,
+      "loss": 1.5185,
       "step": 804
     },
     {
       "epoch": 1.288,
-      "grad_norm": 1.3136061429977417,
-      "learning_rate": 0.00031839999999999993,
-      "loss": 1.6089,
+      "grad_norm": 0.9999681115150452,
+      "learning_rate": 0.0002385,
+      "loss": 1.8636,
       "step": 805
     },
     {
       "epoch": 1.2896,
-      "grad_norm": 1.137098789215088,
-      "learning_rate": 0.00031879999999999994,
-      "loss": 1.5119,
+      "grad_norm": 1.317078709602356,
+      "learning_rate": 0.0002388,
+      "loss": 1.3411,
       "step": 806
     },
     {
       "epoch": 1.2912,
-      "grad_norm": 0.9974707365036011,
-      "learning_rate": 0.0003192,
-      "loss": 1.4021,
+      "grad_norm": 1.2333732843399048,
+      "learning_rate": 0.00023909999999999998,
+      "loss": 1.694,
       "step": 807
     },
     {
       "epoch": 1.2928,
-      "grad_norm": 1.6253764629364014,
-      "learning_rate": 0.00031959999999999996,
-      "loss": 1.6299,
+      "grad_norm": 1.1172211170196533,
+      "learning_rate": 0.0002394,
+      "loss": 1.518,
       "step": 808
     },
     {
       "epoch": 1.2944,
-      "grad_norm": 1.1321923732757568,
-      "learning_rate": 0.00031999999999999997,
-      "loss": 1.5327,
+      "grad_norm": 1.1694917678833008,
+      "learning_rate": 0.0002397,
+      "loss": 1.8538,
       "step": 809
     },
     {
       "epoch": 1.296,
-      "grad_norm": 0.8714154362678528,
-      "learning_rate": 0.0003204,
-      "loss": 1.6207,
+      "grad_norm": 2.515814781188965,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 1.8284,
       "step": 810
     },
     {
       "epoch": 1.2976,
-      "grad_norm": 1.2042042016983032,
-      "learning_rate": 0.00032079999999999994,
-      "loss": 1.6473,
+      "grad_norm": 2.0576000213623047,
+      "learning_rate": 0.00024029999999999999,
+      "loss": 1.9621,
       "step": 811
     },
     {
       "epoch": 1.2992,
-      "grad_norm": 1.5419819355010986,
-      "learning_rate": 0.00032119999999999995,
-      "loss": 1.8554,
+      "grad_norm": 2.3491764068603516,
+      "learning_rate": 0.0002406,
+      "loss": 1.9038,
       "step": 812
     },
     {
       "epoch": 1.3008,
-      "grad_norm": 1.4085866212844849,
-      "learning_rate": 0.0003216,
-      "loss": 1.3106,
+      "grad_norm": 1.76621675491333,
+      "learning_rate": 0.0002409,
+      "loss": 1.637,
       "step": 813
     },
     {
       "epoch": 1.3024,
-      "grad_norm": 1.227091908454895,
-      "learning_rate": 0.00032199999999999997,
-      "loss": 1.7718,
+      "grad_norm": 1.208563208580017,
+      "learning_rate": 0.00024119999999999998,
+      "loss": 1.8757,
       "step": 814
     },
     {
       "epoch": 1.304,
-      "grad_norm": 1.610036015510559,
-      "learning_rate": 0.0003224,
-      "loss": 1.5036,
+      "grad_norm": 1.1317024230957031,
+      "learning_rate": 0.0002415,
+      "loss": 1.8826,
       "step": 815
     },
     {
       "epoch": 1.3056,
-      "grad_norm": 1.597038745880127,
-      "learning_rate": 0.0003228,
-      "loss": 2.0035,
+      "grad_norm": 1.3644945621490479,
+      "learning_rate": 0.0002418,
+      "loss": 1.4626,
       "step": 816
     },
     {
       "epoch": 1.3072,
-      "grad_norm": 1.719772219657898,
-      "learning_rate": 0.00032319999999999994,
-      "loss": 1.7124,
+      "grad_norm": 1.633872628211975,
+      "learning_rate": 0.0002421,
+      "loss": 1.8603,
       "step": 817
     },
     {
       "epoch": 1.3088,
-      "grad_norm": 1.2312437295913696,
-      "learning_rate": 0.00032359999999999995,
-      "loss": 1.664,
+      "grad_norm": 1.3554219007492065,
+      "learning_rate": 0.00024239999999999998,
+      "loss": 1.8397,
       "step": 818
     },
     {
       "epoch": 1.3104,
-      "grad_norm": 1.160129427909851,
-      "learning_rate": 0.000324,
-      "loss": 1.4447,
+      "grad_norm": 1.2498072385787964,
+      "learning_rate": 0.0002427,
+      "loss": 1.8116,
       "step": 819
     },
     {
       "epoch": 1.312,
-      "grad_norm": 1.1790233850479126,
-      "learning_rate": 0.00032439999999999997,
-      "loss": 1.7727,
+      "grad_norm": 1.3917790651321411,
+      "learning_rate": 0.000243,
+      "loss": 1.7565,
       "step": 820
     },
     {
       "epoch": 1.3136,
-      "grad_norm": 1.2099790573120117,
-      "learning_rate": 0.0003248,
-      "loss": 1.6751,
+      "grad_norm": 1.4509116411209106,
+      "learning_rate": 0.0002433,
+      "loss": 1.7601,
       "step": 821
     },
     {
       "epoch": 1.3152,
-      "grad_norm": 1.0591086149215698,
-      "learning_rate": 0.0003252,
-      "loss": 1.6114,
+      "grad_norm": 1.7720223665237427,
+      "learning_rate": 0.00024359999999999999,
+      "loss": 1.6885,
       "step": 822
     },
     {
       "epoch": 1.3168,
-      "grad_norm": 2.331596612930298,
-      "learning_rate": 0.00032559999999999995,
-      "loss": 2.0011,
+      "grad_norm": 1.7985706329345703,
+      "learning_rate": 0.00024389999999999997,
+      "loss": 2.0853,
       "step": 823
     },
     {
       "epoch": 1.3184,
-      "grad_norm": 1.589240312576294,
-      "learning_rate": 0.00032599999999999996,
-      "loss": 1.6671,
+      "grad_norm": 1.8156722784042358,
+      "learning_rate": 0.00024419999999999997,
+      "loss": 2.2531,
       "step": 824
     },
     {
       "epoch": 1.32,
-      "grad_norm": NaN,
-      "learning_rate": 0.00032599999999999996,
-      "loss": 1.9173,
+      "grad_norm": 1.7719063758850098,
+      "learning_rate": 0.0002445,
+      "loss": 2.3759,
       "step": 825
     },
     {
       "epoch": 1.3216,
-      "grad_norm": 6.769994735717773,
-      "learning_rate": 0.0003264,
-      "loss": 2.5387,
+      "grad_norm": 27.36151695251465,
+      "learning_rate": 0.0002448,
+      "loss": 4.4413,
       "step": 826
     },
     {
       "epoch": 1.3232,
-      "grad_norm": 5.106613636016846,
-      "learning_rate": 0.0003268,
-      "loss": 1.9806,
+      "grad_norm": 8.232678413391113,
+      "learning_rate": 0.00024509999999999994,
+      "loss": 2.624,
       "step": 827
     },
     {
       "epoch": 1.3248,
-      "grad_norm": 2.9851677417755127,
-      "learning_rate": 0.0003272,
-      "loss": 1.7926,
+      "grad_norm": 4.0654497146606445,
+      "learning_rate": 0.00024539999999999995,
+      "loss": 1.9936,
       "step": 828
     },
     {
       "epoch": 1.3264,
-      "grad_norm": 1.8340100049972534,
-      "learning_rate": 0.0003276,
-      "loss": 1.9116,
+      "grad_norm": 5.744853973388672,
+      "learning_rate": 0.00024569999999999995,
+      "loss": 2.2401,
       "step": 829
     },
     {
       "epoch": 1.328,
-      "grad_norm": 8.570950508117676,
-      "learning_rate": 0.00032799999999999995,
-      "loss": 2.8044,
+      "grad_norm": 2.1282172203063965,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 1.9586,
       "step": 830
     },
     {
       "epoch": 1.3296000000000001,
-      "grad_norm": 2.8660714626312256,
-      "learning_rate": 0.00032839999999999996,
-      "loss": 1.6779,
+      "grad_norm": 1.5827854871749878,
+      "learning_rate": 0.00024629999999999997,
+      "loss": 1.8686,
       "step": 831
     },
     {
       "epoch": 1.3312,
-      "grad_norm": 1.373335838317871,
-      "learning_rate": 0.0003288,
-      "loss": 2.183,
+      "grad_norm": 3.850166082382202,
+      "learning_rate": 0.0002466,
+      "loss": 1.7661,
       "step": 832
     },
     {
       "epoch": 1.3328,
-      "grad_norm": 2.5283563137054443,
-      "learning_rate": 0.0003292,
-      "loss": 1.6865,
+      "grad_norm": 4.2956719398498535,
+      "learning_rate": 0.0002469,
+      "loss": 1.774,
       "step": 833
     },
     {
       "epoch": 1.3344,
-      "grad_norm": 3.3186419010162354,
-      "learning_rate": 0.0003296,
-      "loss": 1.602,
+      "grad_norm": 3.491035223007202,
+      "learning_rate": 0.0002472,
+      "loss": 1.9191,
       "step": 834
     },
     {
       "epoch": 1.336,
-      "grad_norm": 2.754730463027954,
-      "learning_rate": 0.00033,
-      "loss": 1.4721,
+      "grad_norm": 3.0301005840301514,
+      "learning_rate": 0.00024749999999999994,
+      "loss": 1.8073,
       "step": 835
     },
     {
       "epoch": 1.3376000000000001,
-      "grad_norm": 0.960064172744751,
-      "learning_rate": 0.00033039999999999995,
-      "loss": 1.807,
+      "grad_norm": 1.9792249202728271,
+      "learning_rate": 0.00024779999999999995,
+      "loss": 1.7315,
       "step": 836
     },
     {
       "epoch": 1.3392,
-      "grad_norm": 1.207796573638916,
-      "learning_rate": 0.00033079999999999996,
-      "loss": 1.5925,
+      "grad_norm": 1.0492668151855469,
+      "learning_rate": 0.00024809999999999996,
+      "loss": 1.6863,
       "step": 837
     },
     {
       "epoch": 1.3408,
-      "grad_norm": 2.0469510555267334,
-      "learning_rate": 0.0003312,
-      "loss": 1.6148,
+      "grad_norm": 1.034518837928772,
+      "learning_rate": 0.00024839999999999997,
+      "loss": 1.5301,
       "step": 838
     },
     {
       "epoch": 1.3424,
-      "grad_norm": 1.0148212909698486,
-      "learning_rate": 0.00033159999999999993,
-      "loss": 1.4489,
+      "grad_norm": 0.9527673125267029,
+      "learning_rate": 0.0002487,
+      "loss": 1.509,
       "step": 839
     },
     {
       "epoch": 1.3439999999999999,
-      "grad_norm": 0.936536431312561,
-      "learning_rate": 0.000332,
-      "loss": 1.1779,
+      "grad_norm": 2.35052227973938,
+      "learning_rate": 0.000249,
+      "loss": 1.6575,
       "step": 840
     },
     {
       "epoch": 1.3456000000000001,
-      "grad_norm": 0.9345409274101257,
-      "learning_rate": 0.0003324,
-      "loss": 1.4338,
+      "grad_norm": 1.5617680549621582,
+      "learning_rate": 0.0002493,
+      "loss": 1.3983,
       "step": 841
     },
     {
       "epoch": 1.3472,
-      "grad_norm": 0.8234386444091797,
-      "learning_rate": 0.00033279999999999996,
-      "loss": 1.4326,
+      "grad_norm": 3.1023738384246826,
+      "learning_rate": 0.00024959999999999994,
+      "loss": 1.7373,
       "step": 842
     },
     {
       "epoch": 1.3488,
-      "grad_norm": 1.0428308248519897,
-      "learning_rate": 0.00033319999999999997,
-      "loss": 1.6166,
+      "grad_norm": 1.3363111019134521,
+      "learning_rate": 0.00024989999999999995,
+      "loss": 1.6505,
       "step": 843
     },
     {
       "epoch": 1.3504,
-      "grad_norm": 0.8084268569946289,
-      "learning_rate": 0.0003336,
-      "loss": 1.6855,
+      "grad_norm": 5.273314476013184,
+      "learning_rate": 0.00025019999999999996,
+      "loss": 2.1203,
       "step": 844
     },
     {
       "epoch": 1.3519999999999999,
-      "grad_norm": 1.372387409210205,
-      "learning_rate": 0.00033399999999999993,
-      "loss": 1.5671,
+      "grad_norm": 1.881166934967041,
+      "learning_rate": 0.00025049999999999996,
+      "loss": 1.5656,
       "step": 845
     },
     {
       "epoch": 1.3536000000000001,
-      "grad_norm": 1.301967978477478,
-      "learning_rate": 0.0003344,
-      "loss": 1.4755,
+      "grad_norm": 1.683435082435608,
+      "learning_rate": 0.00025079999999999997,
+      "loss": 1.6089,
       "step": 846
     },
     {
       "epoch": 1.3552,
-      "grad_norm": 0.9071243405342102,
-      "learning_rate": 0.0003348,
-      "loss": 1.3501,
+      "grad_norm": 1.2385189533233643,
+      "learning_rate": 0.0002511,
+      "loss": 1.1731,
       "step": 847
     },
     {
       "epoch": 1.3568,
-      "grad_norm": 1.210505723953247,
-      "learning_rate": 0.00033519999999999996,
-      "loss": 1.5101,
+      "grad_norm": 0.8636699318885803,
+      "learning_rate": 0.0002514,
+      "loss": 1.5439,
       "step": 848
     },
     {
       "epoch": 1.3584,
-      "grad_norm": 2.225215435028076,
-      "learning_rate": 0.00033559999999999997,
-      "loss": 1.6298,
+      "grad_norm": 0.9964292049407959,
+      "learning_rate": 0.0002517,
+      "loss": 1.7566,
       "step": 849
     },
     {
       "epoch": 1.3599999999999999,
-      "grad_norm": 1.9273438453674316,
-      "learning_rate": 0.000336,
-      "loss": 1.5265,
+      "grad_norm": 1.041764497756958,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 1.5745,
       "step": 850
     },
     {
       "epoch": 1.3616,
-      "grad_norm": 0.9094347953796387,
-      "learning_rate": 0.00033639999999999994,
-      "loss": 1.2325,
+      "grad_norm": 0.9211375117301941,
+      "learning_rate": 0.00025229999999999995,
+      "loss": 1.4535,
       "step": 851
     },
     {
       "epoch": 1.3632,
-      "grad_norm": 1.0150439739227295,
-      "learning_rate": 0.0003368,
-      "loss": 1.4602,
+      "grad_norm": 0.7638111710548401,
+      "learning_rate": 0.00025259999999999996,
+      "loss": 1.4994,
       "step": 852
     },
     {
       "epoch": 1.3648,
-      "grad_norm": 0.9719175696372986,
-      "learning_rate": 0.0003372,
-      "loss": 1.7831,
+      "grad_norm": 1.4103494882583618,
+      "learning_rate": 0.00025289999999999997,
+      "loss": 1.3608,
       "step": 853
     },
     {
       "epoch": 1.3664,
-      "grad_norm": 1.3457897901535034,
-      "learning_rate": 0.00033759999999999997,
-      "loss": 1.4172,
+      "grad_norm": 0.9239462614059448,
+      "learning_rate": 0.0002532,
+      "loss": 1.65,
       "step": 854
     },
     {
       "epoch": 1.3679999999999999,
-      "grad_norm": 2.2474122047424316,
-      "learning_rate": 0.000338,
-      "loss": 1.7245,
+      "grad_norm": 1.2657464742660522,
+      "learning_rate": 0.0002535,
+      "loss": 1.4439,
       "step": 855
     },
     {
       "epoch": 1.3696,
-      "grad_norm": 1.9882428646087646,
-      "learning_rate": 0.00033839999999999993,
-      "loss": 1.5253,
+      "grad_norm": 1.411150336265564,
+      "learning_rate": 0.0002538,
+      "loss": 1.5659,
       "step": 856
     },
     {
       "epoch": 1.3712,
-      "grad_norm": 1.4229366779327393,
-      "learning_rate": 0.00033879999999999994,
-      "loss": 1.6048,
+      "grad_norm": 1.1029832363128662,
+      "learning_rate": 0.0002541,
+      "loss": 1.5794,
       "step": 857
     },
     {
       "epoch": 1.3728,
-      "grad_norm": 0.8712606430053711,
-      "learning_rate": 0.0003392,
-      "loss": 1.774,
+      "grad_norm": 1.2146772146224976,
+      "learning_rate": 0.00025439999999999995,
+      "loss": 1.8554,
       "step": 858
     },
     {
       "epoch": 1.3744,
-      "grad_norm": 0.8492836952209473,
-      "learning_rate": 0.00033959999999999996,
-      "loss": 1.5786,
+      "grad_norm": 1.0912986993789673,
+      "learning_rate": 0.00025469999999999996,
+      "loss": 1.9646,
       "step": 859
     },
     {
       "epoch": 1.376,
-      "grad_norm": 1.6412835121154785,
-      "learning_rate": 0.00033999999999999997,
-      "loss": 1.3234,
+      "grad_norm": 1.115552306175232,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 1.4906,
       "step": 860
     },
     {
       "epoch": 1.3776,
-      "grad_norm": 1.2218254804611206,
-      "learning_rate": 0.0003404,
-      "loss": 1.8886,
+      "grad_norm": 1.5120686292648315,
+      "learning_rate": 0.00025529999999999997,
+      "loss": 1.5043,
       "step": 861
     },
     {
       "epoch": 1.3792,
-      "grad_norm": 1.142381191253662,
-      "learning_rate": 0.00034079999999999994,
-      "loss": 1.5134,
+      "grad_norm": 2.1887874603271484,
+      "learning_rate": 0.0002556,
+      "loss": 2.1509,
       "step": 862
     },
     {
       "epoch": 1.3808,
-      "grad_norm": 1.4286154508590698,
-      "learning_rate": 0.00034119999999999995,
-      "loss": 1.3333,
+      "grad_norm": 1.7975490093231201,
+      "learning_rate": 0.0002559,
+      "loss": 1.5343,
       "step": 863
     },
     {
       "epoch": 1.3824,
-      "grad_norm": 0.9015567302703857,
-      "learning_rate": 0.0003416,
-      "loss": 1.5008,
+      "grad_norm": 0.8767451047897339,
+      "learning_rate": 0.0002562,
+      "loss": 1.8846,
       "step": 864
     },
     {
       "epoch": 1.384,
-      "grad_norm": 0.8696253895759583,
-      "learning_rate": 0.00034199999999999996,
-      "loss": 1.8428,
+      "grad_norm": 1.3470810651779175,
+      "learning_rate": 0.00025649999999999995,
+      "loss": 1.8476,
       "step": 865
     },
     {
       "epoch": 1.3856,
-      "grad_norm": 1.8368676900863647,
-      "learning_rate": 0.0003424,
-      "loss": 1.4036,
+      "grad_norm": 1.2314963340759277,
+      "learning_rate": 0.00025679999999999995,
+      "loss": 1.616,
       "step": 866
     },
     {
       "epoch": 1.3872,
-      "grad_norm": 2.2873785495758057,
-      "learning_rate": 0.0003428,
-      "loss": 1.913,
+      "grad_norm": 1.2231523990631104,
+      "learning_rate": 0.00025709999999999996,
+      "loss": 1.7002,
       "step": 867
     },
     {
       "epoch": 1.3888,
-      "grad_norm": 2.610448122024536,
-      "learning_rate": 0.00034319999999999994,
-      "loss": 2.3383,
+      "grad_norm": 1.706369161605835,
+      "learning_rate": 0.00025739999999999997,
+      "loss": 1.882,
       "step": 868
     },
     {
       "epoch": 1.3904,
-      "grad_norm": 1.2571147680282593,
-      "learning_rate": 0.00034359999999999995,
-      "loss": 1.462,
+      "grad_norm": 1.3242183923721313,
+      "learning_rate": 0.0002577,
+      "loss": 1.9856,
       "step": 869
     },
     {
       "epoch": 1.392,
-      "grad_norm": 1.162726640701294,
-      "learning_rate": 0.000344,
-      "loss": 1.4693,
+      "grad_norm": 1.4956287145614624,
+      "learning_rate": 0.000258,
+      "loss": 1.8133,
       "step": 870
     },
     {
       "epoch": 1.3936,
-      "grad_norm": 1.0974887609481812,
-      "learning_rate": 0.00034439999999999997,
-      "loss": 1.7837,
+      "grad_norm": 1.1745412349700928,
+      "learning_rate": 0.0002583,
+      "loss": 2.0496,
       "step": 871
     },
     {
       "epoch": 1.3952,
-      "grad_norm": 6.15885591506958,
-      "learning_rate": 0.0003448,
-      "loss": 2.0816,
+      "grad_norm": 1.716749906539917,
+      "learning_rate": 0.0002586,
+      "loss": 1.8053,
       "step": 872
     },
     {
       "epoch": 1.3968,
-      "grad_norm": 2.189544200897217,
-      "learning_rate": 0.0003452,
-      "loss": 1.6858,
+      "grad_norm": 1.0582464933395386,
+      "learning_rate": 0.00025889999999999995,
+      "loss": 1.8637,
       "step": 873
     },
     {
       "epoch": 1.3984,
-      "grad_norm": 1.991902470588684,
-      "learning_rate": 0.00034559999999999994,
-      "loss": 1.9152,
+      "grad_norm": 1.3968075513839722,
+      "learning_rate": 0.00025919999999999996,
+      "loss": 2.2903,
       "step": 874
     },
     {
       "epoch": 1.4,
-      "grad_norm": 3.02516770362854,
-      "learning_rate": 0.00034599999999999995,
-      "loss": 2.3831,
+      "grad_norm": 1.8292760848999023,
+      "learning_rate": 0.00025949999999999997,
+      "loss": 2.5341,
       "step": 875
     },
     {
       "epoch": 1.4016,
-      "grad_norm": 10.343817710876465,
-      "learning_rate": 0.00034639999999999996,
-      "loss": 2.8638,
+      "grad_norm": 7.413393497467041,
+      "learning_rate": 0.00025979999999999997,
+      "loss": 2.3879,
       "step": 876
     },
     {
       "epoch": 1.4032,
-      "grad_norm": 7.452947616577148,
-      "learning_rate": 0.0003467999999999999,
-      "loss": 2.2708,
+      "grad_norm": 6.438076496124268,
+      "learning_rate": 0.0002601,
+      "loss": 2.4021,
       "step": 877
     },
     {
       "epoch": 1.4048,
-      "grad_norm": 2.4687275886535645,
-      "learning_rate": 0.0003472,
-      "loss": 1.8979,
+      "grad_norm": 1.5802280902862549,
+      "learning_rate": 0.0002604,
+      "loss": 1.7735,
       "step": 878
     },
     {
       "epoch": 1.4064,
-      "grad_norm": 1.3352081775665283,
-      "learning_rate": 0.0003476,
-      "loss": 1.7263,
+      "grad_norm": 3.181586265563965,
+      "learning_rate": 0.0002607,
+      "loss": 2.2812,
       "step": 879
     },
     {
       "epoch": 1.408,
-      "grad_norm": 1.345573902130127,
-      "learning_rate": 0.00034799999999999995,
-      "loss": 2.1026,
+      "grad_norm": 2.4558494091033936,
+      "learning_rate": 0.000261,
+      "loss": 2.1456,
       "step": 880
     },
     {
       "epoch": 1.4096,
-      "grad_norm": 2.9327187538146973,
-      "learning_rate": 0.00034839999999999996,
-      "loss": 2.5618,
+      "grad_norm": 4.251657485961914,
+      "learning_rate": 0.00026129999999999995,
+      "loss": 1.7878,
       "step": 881
     },
     {
       "epoch": 1.4112,
-      "grad_norm": 4.26807975769043,
-      "learning_rate": 0.00034879999999999997,
-      "loss": 1.9103,
+      "grad_norm": 2.2614433765411377,
+      "learning_rate": 0.00026159999999999996,
+      "loss": 2.2472,
       "step": 882
     },
     {
       "epoch": 1.4128,
-      "grad_norm": 5.132489204406738,
-      "learning_rate": 0.0003491999999999999,
-      "loss": 2.1589,
+      "grad_norm": 4.5385966300964355,
+      "learning_rate": 0.00026189999999999997,
+      "loss": 2.0106,
       "step": 883
     },
     {
       "epoch": 1.4144,
-      "grad_norm": 4.745561122894287,
-      "learning_rate": 0.0003496,
-      "loss": 1.7655,
+      "grad_norm": 4.714118957519531,
+      "learning_rate": 0.0002622,
+      "loss": 2.1035,
       "step": 884
     },
     {
       "epoch": 1.416,
-      "grad_norm": 3.5287861824035645,
-      "learning_rate": 0.00035,
-      "loss": 1.661,
+      "grad_norm": 5.232901573181152,
+      "learning_rate": 0.0002625,
+      "loss": 2.0967,
       "step": 885
     },
     {
       "epoch": 1.4176,
-      "grad_norm": 3.1767563819885254,
-      "learning_rate": 0.00035039999999999995,
-      "loss": 1.6753,
+      "grad_norm": 3.031771421432495,
+      "learning_rate": 0.0002628,
+      "loss": 1.8065,
       "step": 886
     },
     {
       "epoch": 1.4192,
-      "grad_norm": 1.0548657178878784,
-      "learning_rate": 0.00035079999999999996,
-      "loss": 1.736,
+      "grad_norm": 2.2014353275299072,
+      "learning_rate": 0.0002631,
+      "loss": 1.7145,
       "step": 887
     },
     {
       "epoch": 1.4208,
-      "grad_norm": 1.0605950355529785,
-      "learning_rate": 0.00035119999999999997,
-      "loss": 1.7661,
+      "grad_norm": 1.0654090642929077,
+      "learning_rate": 0.00026339999999999995,
+      "loss": 1.5715,
       "step": 888
     },
     {
       "epoch": 1.4224,
-      "grad_norm": 2.0522172451019287,
-      "learning_rate": 0.0003515999999999999,
-      "loss": 1.2061,
+      "grad_norm": 1.569884181022644,
+      "learning_rate": 0.00026369999999999996,
+      "loss": 1.4741,
       "step": 889
     },
     {
       "epoch": 1.424,
-      "grad_norm": 3.3747687339782715,
-      "learning_rate": 0.000352,
-      "loss": 1.5711,
+      "grad_norm": 2.636420249938965,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 1.48,
       "step": 890
     },
     {
       "epoch": 1.4256,
-      "grad_norm": 3.890195608139038,
-      "learning_rate": 0.0003524,
-      "loss": 1.9462,
+      "grad_norm": 3.305544376373291,
+      "learning_rate": 0.0002643,
+      "loss": 1.487,
       "step": 891
     },
     {
       "epoch": 1.4272,
-      "grad_norm": 3.8554182052612305,
-      "learning_rate": 0.00035279999999999996,
-      "loss": 1.9326,
+      "grad_norm": 3.3960185050964355,
+      "learning_rate": 0.0002646,
+      "loss": 1.43,
       "step": 892
     },
     {
       "epoch": 1.4288,
-      "grad_norm": 3.2069602012634277,
-      "learning_rate": 0.00035319999999999997,
-      "loss": 1.7907,
+      "grad_norm": 2.4552576541900635,
+      "learning_rate": 0.0002649,
+      "loss": 1.4312,
       "step": 893
     },
     {
       "epoch": 1.4304000000000001,
-      "grad_norm": 3.5373735427856445,
-      "learning_rate": 0.0003536,
-      "loss": 1.6313,
+      "grad_norm": 2.176464557647705,
+      "learning_rate": 0.0002652,
+      "loss": 1.6457,
       "step": 894
     },
     {
       "epoch": 1.432,
-      "grad_norm": 2.3303987979888916,
-      "learning_rate": 0.00035399999999999993,
-      "loss": 1.4145,
+      "grad_norm": 1.3099313974380493,
+      "learning_rate": 0.0002655,
+      "loss": 1.3961,
       "step": 895
     },
     {
       "epoch": 1.4336,
-      "grad_norm": 2.315999984741211,
-      "learning_rate": 0.0003544,
-      "loss": 1.7224,
+      "grad_norm": 1.3967254161834717,
+      "learning_rate": 0.00026579999999999996,
+      "loss": 1.8732,
       "step": 896
     },
     {
       "epoch": 1.4352,
-      "grad_norm": 1.1353230476379395,
-      "learning_rate": 0.0003548,
-      "loss": 1.6141,
+      "grad_norm": 2.6144227981567383,
+      "learning_rate": 0.00026609999999999996,
+      "loss": 1.9415,
       "step": 897
     },
     {
       "epoch": 1.4368,
-      "grad_norm": 1.321402907371521,
-      "learning_rate": 0.00035519999999999996,
-      "loss": 1.5741,
+      "grad_norm": 2.4597859382629395,
+      "learning_rate": 0.00026639999999999997,
+      "loss": 1.6611,
       "step": 898
     },
     {
       "epoch": 1.4384000000000001,
-      "grad_norm": 1.3665454387664795,
-      "learning_rate": 0.00035559999999999997,
-      "loss": 1.3824,
+      "grad_norm": 3.06152081489563,
+      "learning_rate": 0.0002667,
+      "loss": 1.7697,
       "step": 899
     },
     {
       "epoch": 1.44,
-      "grad_norm": 1.6859432458877563,
-      "learning_rate": 0.000356,
-      "loss": 1.6852,
-      "step": 900
-    },
-    {
-      "epoch": 1.44,
-      "eval_cer": 0.4112289233467533,
-      "eval_loss": 1.986970067024231,
-      "eval_runtime": 157.7785,
-      "eval_samples_per_second": 19.876,
-      "eval_steps_per_second": 1.242,
-      "eval_wer": 0.6153973681483784,
+      "grad_norm": 3.207355260848999,
+      "learning_rate": 0.000267,
+      "loss": 1.699,
       "step": 900
     },
     {
       "epoch": 1.4416,
-      "grad_norm": 1.5820800065994263,
-      "learning_rate": 0.00035639999999999994,
-      "loss": 1.4903,
+      "grad_norm": 2.3851301670074463,
+      "learning_rate": 0.0002673,
+      "loss": 2.0775,
       "step": 901
     },
     {
       "epoch": 1.4432,
-      "grad_norm": 1.1886115074157715,
-      "learning_rate": 0.0003568,
-      "loss": 1.6061,
+      "grad_norm": 1.7599847316741943,
+      "learning_rate": 0.0002676,
+      "loss": 1.8226,
       "step": 902
     },
     {
       "epoch": 1.4447999999999999,
-      "grad_norm": 1.2537235021591187,
-      "learning_rate": 0.0003572,
-      "loss": 1.4464,
+      "grad_norm": 0.9662289023399353,
+      "learning_rate": 0.0002679,
+      "loss": 1.7798,
       "step": 903
     },
     {
       "epoch": 1.4464000000000001,
-      "grad_norm": 1.027907133102417,
-      "learning_rate": 0.00035759999999999996,
-      "loss": 1.2771,
+      "grad_norm": 0.8772083520889282,
+      "learning_rate": 0.00026819999999999996,
+      "loss": 1.4852,
       "step": 904
     },
     {
       "epoch": 1.448,
-      "grad_norm": 1.5466809272766113,
-      "learning_rate": 0.000358,
-      "loss": 1.3082,
+      "grad_norm": 2.0723578929901123,
+      "learning_rate": 0.00026849999999999997,
+      "loss": 1.5108,
       "step": 905
     },
     {
       "epoch": 1.4496,
-      "grad_norm": 1.8753767013549805,
-      "learning_rate": 0.0003584,
-      "loss": 1.2227,
+      "grad_norm": 2.259298086166382,
+      "learning_rate": 0.0002688,
+      "loss": 1.6322,
       "step": 906
     },
     {
       "epoch": 1.4512,
-      "grad_norm": 2.8376686573028564,
-      "learning_rate": 0.00035879999999999994,
-      "loss": 1.8457,
+      "grad_norm": 1.930001974105835,
+      "learning_rate": 0.0002691,
+      "loss": 1.834,
       "step": 907
     },
     {
       "epoch": 1.4527999999999999,
-      "grad_norm": 2.1180849075317383,
-      "learning_rate": 0.0003592,
-      "loss": 1.5696,
+      "grad_norm": 1.9969251155853271,
+      "learning_rate": 0.0002694,
+      "loss": 1.4405,
       "step": 908
     },
     {
       "epoch": 1.4544000000000001,
-      "grad_norm": 1.1068503856658936,
-      "learning_rate": 0.0003596,
-      "loss": 1.484,
+      "grad_norm": 2.078216552734375,
+      "learning_rate": 0.0002697,
+      "loss": 2.1698,
       "step": 909
     },
     {
       "epoch": 1.456,
-      "grad_norm": 1.0974130630493164,
-      "learning_rate": 0.00035999999999999997,
-      "loss": 1.6244,
+      "grad_norm": 1.2375624179840088,
+      "learning_rate": 0.00027,
+      "loss": 1.7028,
       "step": 910
     },
     {
       "epoch": 1.4576,
-      "grad_norm": 1.0330309867858887,
-      "learning_rate": 0.0003604,
-      "loss": 1.8489,
+      "grad_norm": 1.521498203277588,
+      "learning_rate": 0.00027029999999999996,
+      "loss": 1.7956,
       "step": 911
     },
     {
       "epoch": 1.4592,
-      "grad_norm": 1.4640865325927734,
-      "learning_rate": 0.0003608,
-      "loss": 1.8396,
+      "grad_norm": 1.4378412961959839,
+      "learning_rate": 0.00027059999999999996,
+      "loss": 1.7424,
       "step": 912
     },
     {
       "epoch": 1.4607999999999999,
-      "grad_norm": 1.0381420850753784,
-      "learning_rate": 0.00036119999999999994,
-      "loss": 1.7392,
+      "grad_norm": 1.0808361768722534,
+      "learning_rate": 0.00027089999999999997,
+      "loss": 1.4528,
       "step": 913
     },
     {
       "epoch": 1.4624,
-      "grad_norm": 0.9180180430412292,
-      "learning_rate": 0.00036159999999999995,
-      "loss": 1.5514,
+      "grad_norm": 1.5603634119033813,
+      "learning_rate": 0.0002712,
+      "loss": 2.0128,
       "step": 914
     },
     {
       "epoch": 1.464,
-      "grad_norm": 1.0205705165863037,
-      "learning_rate": 0.000362,
-      "loss": 1.5544,
+      "grad_norm": 1.2612849473953247,
+      "learning_rate": 0.0002715,
+      "loss": 1.7135,
       "step": 915
     },
     {
       "epoch": 1.4656,
-      "grad_norm": 0.8514875769615173,
-      "learning_rate": 0.00036239999999999997,
-      "loss": 1.6248,
+      "grad_norm": 1.4489110708236694,
+      "learning_rate": 0.0002718,
+      "loss": 1.5092,
       "step": 916
     },
     {
       "epoch": 1.4672,
-      "grad_norm": 1.1364916563034058,
-      "learning_rate": 0.0003628,
-      "loss": 1.5353,
+      "grad_norm": 1.5542961359024048,
+      "learning_rate": 0.0002721,
+      "loss": 1.5612,
       "step": 917
     },
     {
       "epoch": 1.4687999999999999,
-      "grad_norm": 1.7738701105117798,
-      "learning_rate": 0.00036319999999999994,
-      "loss": 1.6905,
+      "grad_norm": 1.411827802658081,
+      "learning_rate": 0.0002724,
+      "loss": 1.5166,
       "step": 918
     },
     {
       "epoch": 1.4704,
-      "grad_norm": 3.32802414894104,
-      "learning_rate": 0.00036359999999999995,
-      "loss": 2.08,
+      "grad_norm": 1.3303709030151367,
+      "learning_rate": 0.00027269999999999996,
+      "loss": 1.3164,
       "step": 919
     },
     {
       "epoch": 1.472,
-      "grad_norm": 1.756226658821106,
-      "learning_rate": 0.00036399999999999996,
-      "loss": 1.6969,
+      "grad_norm": 1.1644420623779297,
+      "learning_rate": 0.00027299999999999997,
+      "loss": 1.8777,
       "step": 920
     },
     {
       "epoch": 1.4736,
-      "grad_norm": 1.2360084056854248,
-      "learning_rate": 0.0003643999999999999,
-      "loss": 1.4355,
+      "grad_norm": 1.7243609428405762,
+      "learning_rate": 0.0002733,
+      "loss": 2.0992,
       "step": 921
     },
     {
       "epoch": 1.4752,
-      "grad_norm": 1.6236251592636108,
-      "learning_rate": 0.0003648,
-      "loss": 2.381,
+      "grad_norm": 1.2297067642211914,
+      "learning_rate": 0.0002736,
+      "loss": 1.5551,
       "step": 922
     },
     {
       "epoch": 1.4768,
-      "grad_norm": 1.5009859800338745,
-      "learning_rate": 0.0003652,
-      "loss": 1.8531,
+      "grad_norm": 1.3536406755447388,
+      "learning_rate": 0.0002739,
+      "loss": 2.1109,
       "step": 923
     },
     {
       "epoch": 1.4784,
-      "grad_norm": 1.9892377853393555,
-      "learning_rate": 0.00036559999999999994,
-      "loss": 2.021,
+      "grad_norm": 1.1590725183486938,
+      "learning_rate": 0.0002742,
+      "loss": 1.965,
       "step": 924
     },
     {
       "epoch": 1.48,
-      "grad_norm": NaN,
-      "learning_rate": 0.00036559999999999994,
-      "loss": 1.6185,
+      "grad_norm": 2.9230568408966064,
+      "learning_rate": 0.0002745,
+      "loss": 2.3223,
       "step": 925
     },
     {
       "epoch": 1.4816,
-      "grad_norm": 8.870469093322754,
-      "learning_rate": 0.00036599999999999995,
-      "loss": 2.6913,
+      "grad_norm": 5.955326557159424,
+      "learning_rate": 0.0002748,
+      "loss": 2.0155,
       "step": 926
     },
     {
       "epoch": 1.4832,
-      "grad_norm": 6.322633743286133,
-      "learning_rate": 0.00036639999999999996,
-      "loss": 2.6344,
+      "grad_norm": 5.281221866607666,
+      "learning_rate": 0.00027509999999999996,
+      "loss": 2.0306,
       "step": 927
     },
     {
       "epoch": 1.4848,
-      "grad_norm": 3.020659923553467,
-      "learning_rate": 0.0003667999999999999,
-      "loss": 1.8173,
+      "grad_norm": 1.9012714624404907,
+      "learning_rate": 0.00027539999999999997,
+      "loss": 1.6593,
       "step": 928
     },
     {
       "epoch": 1.4864,
-      "grad_norm": 2.220780372619629,
-      "learning_rate": 0.0003672,
-      "loss": 1.9342,
+      "grad_norm": 1.502999186515808,
+      "learning_rate": 0.0002757,
+      "loss": 2.0885,
       "step": 929
     },
     {
       "epoch": 1.488,
-      "grad_norm": 1.8951246738433838,
-      "learning_rate": 0.0003676,
-      "loss": 1.6929,
+      "grad_norm": 2.206328868865967,
+      "learning_rate": 0.000276,
+      "loss": 1.5398,
       "step": 930
     },
     {
       "epoch": 1.4896,
-      "grad_norm": 1.6866941452026367,
-      "learning_rate": 0.00036799999999999995,
-      "loss": 2.0583,
+      "grad_norm": 2.663071393966675,
+      "learning_rate": 0.0002763,
+      "loss": 1.6801,
       "step": 931
     },
     {
       "epoch": 1.4912,
-      "grad_norm": 2.816363573074341,
-      "learning_rate": 0.00036839999999999996,
-      "loss": 1.6559,
+      "grad_norm": 2.390557050704956,
+      "learning_rate": 0.0002766,
+      "loss": 1.6475,
       "step": 932
     },
     {
       "epoch": 1.4928,
-      "grad_norm": 1.5674623250961304,
-      "learning_rate": 0.00036879999999999997,
-      "loss": 2.0277,
+      "grad_norm": 1.131935954093933,
+      "learning_rate": 0.0002769,
+      "loss": 1.7341,
       "step": 933
     },
     {
       "epoch": 1.4944,
-      "grad_norm": 3.0781469345092773,
-      "learning_rate": 0.0003691999999999999,
-      "loss": 1.769,
+      "grad_norm": 1.287434458732605,
+      "learning_rate": 0.0002772,
+      "loss": 1.7195,
       "step": 934
     },
     {
       "epoch": 1.496,
-      "grad_norm": 1.0538510084152222,
-      "learning_rate": 0.0003696,
-      "loss": 1.4605,
+      "grad_norm": 0.8551505208015442,
+      "learning_rate": 0.00027749999999999997,
+      "loss": 1.5404,
       "step": 935
     },
     {
       "epoch": 1.4976,
-      "grad_norm": 0.8226996660232544,
-      "learning_rate": 0.00037,
-      "loss": 1.7031,
+      "grad_norm": 1.8731048107147217,
+      "learning_rate": 0.0002778,
+      "loss": 1.495,
       "step": 936
     },
     {
       "epoch": 1.4992,
-      "grad_norm": 3.8168036937713623,
-      "learning_rate": 0.00037039999999999995,
+      "grad_norm": 3.6576366424560547,
+      "learning_rate": 0.0002781,
       "loss": 1.7541,
       "step": 937
     },
     {
       "epoch": 1.5008,
-      "grad_norm": 2.1976897716522217,
-      "learning_rate": 0.00037079999999999996,
-      "loss": 1.5996,
+      "grad_norm": 1.2283856868743896,
+      "learning_rate": 0.0002784,
+      "loss": 1.4094,
       "step": 938
     },
     {
       "epoch": 1.5024,
-      "grad_norm": 1.8852084875106812,
-      "learning_rate": 0.00037119999999999997,
-      "loss": 1.8051,
+      "grad_norm": 2.0950043201446533,
+      "learning_rate": 0.0002787,
+      "loss": 1.3984,
       "step": 939
     },
     {
       "epoch": 1.504,
-      "grad_norm": 1.8353817462921143,
-      "learning_rate": 0.0003715999999999999,
-      "loss": 1.2877,
+      "grad_norm": 1.1806004047393799,
+      "learning_rate": 0.000279,
+      "loss": 1.4307,
       "step": 940
     },
     {
       "epoch": 1.5056,
-      "grad_norm": 1.211934208869934,
-      "learning_rate": 0.000372,
-      "loss": 1.3873,
+      "grad_norm": 0.9991218447685242,
+      "learning_rate": 0.0002793,
+      "loss": 1.2988,
       "step": 941
     },
     {
       "epoch": 1.5072,
-      "grad_norm": 0.9736316204071045,
-      "learning_rate": 0.0003724,
-      "loss": 1.4598,
+      "grad_norm": 1.1656785011291504,
+      "learning_rate": 0.00027959999999999997,
+      "loss": 1.4822,
       "step": 942
     },
     {
       "epoch": 1.5088,
-      "grad_norm": 0.7427512407302856,
-      "learning_rate": 0.00037279999999999995,
-      "loss": 1.3927,
+      "grad_norm": 1.2371271848678589,
+      "learning_rate": 0.0002799,
+      "loss": 1.6206,
       "step": 943
     },
     {
       "epoch": 1.5104,
-      "grad_norm": 1.527969479560852,
-      "learning_rate": 0.00037319999999999996,
-      "loss": 1.384,
+      "grad_norm": 1.1998454332351685,
+      "learning_rate": 0.0002802,
+      "loss": 1.4094,
       "step": 944
     },
     {
       "epoch": 1.512,
-      "grad_norm": 1.2148116827011108,
-      "learning_rate": 0.0003736,
-      "loss": 1.4868,
+      "grad_norm": 1.6202266216278076,
+      "learning_rate": 0.0002805,
+      "loss": 1.6976,
       "step": 945
     },
     {
       "epoch": 1.5135999999999998,
-      "grad_norm": 0.6790880560874939,
-      "learning_rate": 0.00037399999999999993,
-      "loss": 1.3247,
+      "grad_norm": 0.9059401750564575,
+      "learning_rate": 0.0002808,
+      "loss": 1.4302,
       "step": 946
     },
     {
       "epoch": 1.5152,
-      "grad_norm": 0.8779526948928833,
-      "learning_rate": 0.0003744,
-      "loss": 1.5266,
+      "grad_norm": 1.1870366334915161,
+      "learning_rate": 0.0002811,
+      "loss": 1.3084,
       "step": 947
     },
     {
       "epoch": 1.5168,
-      "grad_norm": 1.2756925821304321,
-      "learning_rate": 0.0003748,
-      "loss": 1.3128,
+      "grad_norm": 0.9611499309539795,
+      "learning_rate": 0.00028139999999999996,
+      "loss": 1.2804,
       "step": 948
     },
     {
       "epoch": 1.5184,
-      "grad_norm": 4.33424186706543,
-      "learning_rate": 0.00037519999999999996,
-      "loss": 1.7496,
+      "grad_norm": 1.8231072425842285,
+      "learning_rate": 0.00028169999999999996,
+      "loss": 1.9321,
       "step": 949
     },
     {
       "epoch": 1.52,
-      "grad_norm": 0.9938877820968628,
-      "learning_rate": 0.00037559999999999997,
-      "loss": 1.7513,
+      "grad_norm": 1.3906331062316895,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 1.5093,
       "step": 950
     },
     {
       "epoch": 1.5215999999999998,
-      "grad_norm": 1.0021151304244995,
-      "learning_rate": 0.000376,
-      "loss": 1.4579,
+      "grad_norm": 0.8848568201065063,
+      "learning_rate": 0.0002823,
+      "loss": 1.4336,
       "step": 951
     },
     {
       "epoch": 1.5232,
-      "grad_norm": 1.2110190391540527,
-      "learning_rate": 0.00037639999999999993,
-      "loss": 1.4992,
+      "grad_norm": 0.9558936953544617,
+      "learning_rate": 0.0002826,
+      "loss": 1.6246,
       "step": 952
     },
     {
       "epoch": 1.5248,
-      "grad_norm": 1.012303113937378,
-      "learning_rate": 0.00037679999999999994,
-      "loss": 1.4858,
+      "grad_norm": 0.9805288314819336,
+      "learning_rate": 0.00028289999999999994,
+      "loss": 1.3846,
       "step": 953
     },
     {
       "epoch": 1.5264,
-      "grad_norm": 0.963055431842804,
-      "learning_rate": 0.0003772,
-      "loss": 1.2367,
+      "grad_norm": 0.8371070623397827,
+      "learning_rate": 0.00028319999999999994,
+      "loss": 1.7919,
       "step": 954
     },
     {
       "epoch": 1.528,
-      "grad_norm": 1.213270664215088,
-      "learning_rate": 0.00037759999999999996,
-      "loss": 1.4012,
+      "grad_norm": 0.9453789591789246,
+      "learning_rate": 0.00028349999999999995,
+      "loss": 1.3124,
       "step": 955
     },
     {
       "epoch": 1.5295999999999998,
-      "grad_norm": 1.1823625564575195,
-      "learning_rate": 0.00037799999999999997,
-      "loss": 1.5314,
+      "grad_norm": 1.1171185970306396,
+      "learning_rate": 0.00028379999999999996,
+      "loss": 1.5918,
       "step": 956
     },
     {
       "epoch": 1.5312000000000001,
-      "grad_norm": 1.8237582445144653,
-      "learning_rate": 0.0003784,
-      "loss": 1.5552,
+      "grad_norm": 1.1895886659622192,
+      "learning_rate": 0.00028409999999999997,
+      "loss": 1.3716,
       "step": 957
     },
     {
       "epoch": 1.5328,
-      "grad_norm": 1.1007258892059326,
-      "learning_rate": 0.00037879999999999994,
-      "loss": 1.3313,
+      "grad_norm": 0.9379684329032898,
+      "learning_rate": 0.0002844,
+      "loss": 1.453,
       "step": 958
     },
     {
       "epoch": 1.5344,
-      "grad_norm": 1.2367926836013794,
-      "learning_rate": 0.00037919999999999995,
-      "loss": 1.8743,
+      "grad_norm": 1.0332472324371338,
+      "learning_rate": 0.0002847,
+      "loss": 1.5237,
       "step": 959
     },
     {
       "epoch": 1.536,
-      "grad_norm": 0.8619275689125061,
-      "learning_rate": 0.0003796,
-      "loss": 1.6997,
+      "grad_norm": 0.8858595490455627,
+      "learning_rate": 0.000285,
+      "loss": 1.6488,
       "step": 960
     },
     {
       "epoch": 1.5375999999999999,
-      "grad_norm": 1.0113097429275513,
-      "learning_rate": 0.00037999999999999997,
-      "loss": 1.4948,
+      "grad_norm": 1.681465983390808,
+      "learning_rate": 0.00028529999999999994,
+      "loss": 1.6078,
       "step": 961
     },
     {
       "epoch": 1.5392000000000001,
-      "grad_norm": 1.1451879739761353,
-      "learning_rate": 0.0003804,
-      "loss": 1.7052,
+      "grad_norm": 1.2469699382781982,
+      "learning_rate": 0.00028559999999999995,
+      "loss": 1.7671,
       "step": 962
     },
     {
       "epoch": 1.5408,
-      "grad_norm": 0.9698196053504944,
-      "learning_rate": 0.0003808,
-      "loss": 1.1601,
+      "grad_norm": 1.0610872507095337,
+      "learning_rate": 0.00028589999999999996,
+      "loss": 1.3722,
       "step": 963
     },
     {
       "epoch": 1.5424,
-      "grad_norm": 0.9092131853103638,
-      "learning_rate": 0.00038119999999999994,
-      "loss": 1.7966,
+      "grad_norm": 1.0731545686721802,
+      "learning_rate": 0.00028619999999999996,
+      "loss": 1.7164,
       "step": 964
     },
     {
       "epoch": 1.544,
-      "grad_norm": 1.1889277696609497,
-      "learning_rate": 0.00038159999999999995,
-      "loss": 1.8823,
+      "grad_norm": 1.1132243871688843,
+      "learning_rate": 0.00028649999999999997,
+      "loss": 2.0572,
       "step": 965
     },
     {
       "epoch": 1.5455999999999999,
-      "grad_norm": 1.0519229173660278,
-      "learning_rate": 0.000382,
-      "loss": 1.9142,
+      "grad_norm": 1.1051868200302124,
+      "learning_rate": 0.0002868,
+      "loss": 1.6496,
       "step": 966
     },
     {
       "epoch": 1.5472000000000001,
-      "grad_norm": 1.3340773582458496,
-      "learning_rate": 0.00038239999999999997,
-      "loss": 1.9151,
+      "grad_norm": 1.20839262008667,
+      "learning_rate": 0.0002871,
+      "loss": 1.6104,
       "step": 967
     },
     {
       "epoch": 1.5488,
-      "grad_norm": 2.726638078689575,
-      "learning_rate": 0.0003828,
-      "loss": 1.9075,
+      "grad_norm": 1.0541168451309204,
+      "learning_rate": 0.00028739999999999994,
+      "loss": 1.5231,
       "step": 968
     },
     {
       "epoch": 1.5504,
-      "grad_norm": 2.050051212310791,
-      "learning_rate": 0.0003832,
-      "loss": 1.7483,
+      "grad_norm": 2.385976791381836,
+      "learning_rate": 0.00028769999999999995,
+      "loss": 1.5045,
       "step": 969
     },
     {
       "epoch": 1.552,
-      "grad_norm": 1.1615999937057495,
-      "learning_rate": 0.00038359999999999995,
-      "loss": 1.6727,
+      "grad_norm": 1.1696312427520752,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 1.7261,
       "step": 970
     },
     {
       "epoch": 1.5535999999999999,
-      "grad_norm": 1.1875585317611694,
-      "learning_rate": 0.00038399999999999996,
-      "loss": 1.5908,
+      "grad_norm": 2.113093614578247,
+      "learning_rate": 0.00028829999999999996,
+      "loss": 1.9908,
       "step": 971
     },
     {
       "epoch": 1.5552000000000001,
-      "grad_norm": 1.4524774551391602,
-      "learning_rate": 0.0003844,
-      "loss": 2.023,
+      "grad_norm": 1.855879783630371,
+      "learning_rate": 0.00028859999999999997,
+      "loss": 2.2391,
       "step": 972
     },
     {
       "epoch": 1.5568,
-      "grad_norm": 1.8066257238388062,
-      "learning_rate": 0.0003848,
-      "loss": 2.118,
+      "grad_norm": 2.2188405990600586,
+      "learning_rate": 0.0002889,
+      "loss": 2.032,
       "step": 973
     },
     {
       "epoch": 1.5584,
-      "grad_norm": 1.7197728157043457,
-      "learning_rate": 0.0003852,
-      "loss": 2.0651,
+      "grad_norm": 1.8611500263214111,
+      "learning_rate": 0.0002892,
+      "loss": 2.2839,
       "step": 974
     },
     {
       "epoch": 1.56,
-      "grad_norm": 1.7070485353469849,
-      "learning_rate": 0.0003856,
-      "loss": 2.582,
+      "grad_norm": 2.133455991744995,
+      "learning_rate": 0.0002895,
+      "loss": 2.6324,
       "step": 975
     },
     {
       "epoch": 1.5615999999999999,
-      "grad_norm": 7.113646030426025,
-      "learning_rate": 0.00038599999999999995,
-      "loss": 2.124,
+      "grad_norm": 4.41377067565918,
+      "learning_rate": 0.00028979999999999994,
+      "loss": 2.2812,
       "step": 976
     },
     {
       "epoch": 1.5632000000000001,
-      "grad_norm": 9.63016128540039,
-      "learning_rate": 0.00038639999999999996,
-      "loss": 3.0331,
+      "grad_norm": 1.5284154415130615,
+      "learning_rate": 0.00029009999999999995,
+      "loss": 1.8738,
       "step": 977
     },
     {
       "epoch": 1.5648,
-      "grad_norm": 3.375418186187744,
-      "learning_rate": 0.0003868,
-      "loss": 1.7608,
+      "grad_norm": 1.4382925033569336,
+      "learning_rate": 0.00029039999999999996,
+      "loss": 1.7074,
       "step": 978
     },
     {
       "epoch": 1.5664,
-      "grad_norm": 3.5243453979492188,
-      "learning_rate": 0.0003872,
-      "loss": 1.9515,
+      "grad_norm": 2.024181604385376,
+      "learning_rate": 0.00029069999999999996,
+      "loss": 2.068,
       "step": 979
     },
     {
       "epoch": 1.568,
-      "grad_norm": 3.2788093090057373,
-      "learning_rate": 0.0003876,
-      "loss": 1.9329,
+      "grad_norm": 1.7730194330215454,
+      "learning_rate": 0.00029099999999999997,
+      "loss": 1.8904,
       "step": 980
     },
     {
       "epoch": 1.5695999999999999,
-      "grad_norm": 1.3431663513183594,
-      "learning_rate": 0.00038799999999999994,
-      "loss": 1.849,
+      "grad_norm": 2.1946470737457275,
+      "learning_rate": 0.0002913,
+      "loss": 1.8493,
       "step": 981
     },
     {
       "epoch": 1.5712000000000002,
-      "grad_norm": 3.8363196849823,
-      "learning_rate": 0.00038839999999999995,
-      "loss": 1.7101,
+      "grad_norm": 2.3339195251464844,
+      "learning_rate": 0.0002916,
+      "loss": 1.5104,
       "step": 982
     },
     {
       "epoch": 1.5728,
-      "grad_norm": 4.8841705322265625,
-      "learning_rate": 0.00038879999999999996,
-      "loss": 1.9458,
+      "grad_norm": 1.3478624820709229,
+      "learning_rate": 0.0002919,
+      "loss": 1.4885,
       "step": 983
     },
     {
       "epoch": 1.5744,
-      "grad_norm": 4.90728759765625,
-      "learning_rate": 0.0003891999999999999,
-      "loss": 2.0958,
+      "grad_norm": 1.4026468992233276,
+      "learning_rate": 0.00029219999999999995,
+      "loss": 1.6617,
       "step": 984
     },
     {
       "epoch": 1.576,
-      "grad_norm": 4.307504177093506,
-      "learning_rate": 0.0003896,
-      "loss": 1.8693,
+      "grad_norm": 0.9361331462860107,
+      "learning_rate": 0.00029249999999999995,
+      "loss": 1.5275,
       "step": 985
     },
     {
       "epoch": 1.5776,
-      "grad_norm": 2.8895435333251953,
-      "learning_rate": 0.00039,
-      "loss": 1.7851,
+      "grad_norm": 1.9895044565200806,
+      "learning_rate": 0.00029279999999999996,
+      "loss": 1.5585,
       "step": 986
     },
     {
       "epoch": 1.5792000000000002,
-      "grad_norm": 1.0250154733657837,
-      "learning_rate": 0.00039039999999999995,
-      "loss": 1.7486,
+      "grad_norm": 6.917726516723633,
+      "learning_rate": 0.00029309999999999997,
+      "loss": 2.5852,
       "step": 987
     },
     {
       "epoch": 1.5808,
-      "grad_norm": 0.7546022534370422,
-      "learning_rate": 0.00039079999999999996,
-      "loss": 1.2908,
+      "grad_norm": 1.4938806295394897,
+      "learning_rate": 0.0002934,
+      "loss": 1.2511,
       "step": 988
     },
     {
       "epoch": 1.5824,
-      "grad_norm": 1.4536306858062744,
-      "learning_rate": 0.00039119999999999997,
-      "loss": 1.3772,
+      "grad_norm": 1.2143341302871704,
+      "learning_rate": 0.0002937,
+      "loss": 1.4005,
       "step": 989
     },
     {
       "epoch": 1.584,
-      "grad_norm": 3.1315858364105225,
-      "learning_rate": 0.0003915999999999999,
-      "loss": 1.7422,
+      "grad_norm": 1.1358860731124878,
+      "learning_rate": 0.000294,
+      "loss": 1.6835,
       "step": 990
     },
     {
       "epoch": 1.5856,
-      "grad_norm": 3.7343127727508545,
-      "learning_rate": 0.00039199999999999993,
-      "loss": 1.5513,
+      "grad_norm": 1.1960505247116089,
+      "learning_rate": 0.00029429999999999994,
+      "loss": 1.1956,
       "step": 991
     },
     {
       "epoch": 1.5872000000000002,
-      "grad_norm": 2.6926112174987793,
-      "learning_rate": 0.0003924,
-      "loss": 1.538,
+      "grad_norm": 0.7678542137145996,
+      "learning_rate": 0.00029459999999999995,
+      "loss": 1.3352,
       "step": 992
     },
     {
       "epoch": 1.5888,
-      "grad_norm": 2.2124907970428467,
-      "learning_rate": 0.00039279999999999995,
-      "loss": 1.3235,
+      "grad_norm": 0.9173956513404846,
+      "learning_rate": 0.00029489999999999996,
+      "loss": 1.23,
       "step": 993
     },
     {
       "epoch": 1.5904,
-      "grad_norm": 1.3270829916000366,
-      "learning_rate": 0.00039319999999999996,
-      "loss": 1.405,
+      "grad_norm": 0.8378310799598694,
+      "learning_rate": 0.00029519999999999997,
+      "loss": 1.6007,
       "step": 994
     },
     {
       "epoch": 1.592,
-      "grad_norm": 1.0704325437545776,
-      "learning_rate": 0.00039359999999999997,
-      "loss": 1.5179,
+      "grad_norm": 1.1078053712844849,
+      "learning_rate": 0.00029549999999999997,
+      "loss": 1.4643,
       "step": 995
     },
     {
       "epoch": 1.5936,
-      "grad_norm": 0.7627701759338379,
-      "learning_rate": 0.00039399999999999993,
-      "loss": 1.4045,
+      "grad_norm": 1.037882924079895,
+      "learning_rate": 0.0002958,
+      "loss": 1.3761,
       "step": 996
     },
     {
       "epoch": 1.5952,
-      "grad_norm": 1.1814078092575073,
-      "learning_rate": 0.00039439999999999994,
-      "loss": 1.205,
+      "grad_norm": 0.9665654301643372,
+      "learning_rate": 0.0002961,
+      "loss": 1.5703,
       "step": 997
     },
     {
       "epoch": 1.5968,
-      "grad_norm": 1.1949654817581177,
-      "learning_rate": 0.0003948,
-      "loss": 1.5852,
+      "grad_norm": 0.8781237006187439,
+      "learning_rate": 0.0002964,
+      "loss": 1.3703,
       "step": 998
     },
     {
       "epoch": 1.5984,
-      "grad_norm": 1.570246696472168,
-      "learning_rate": 0.00039519999999999996,
-      "loss": 1.4206,
+      "grad_norm": 1.6409205198287964,
+      "learning_rate": 0.00029669999999999995,
+      "loss": 1.3117,
       "step": 999
     },
     {
       "epoch": 1.6,
-      "grad_norm": 1.0637398958206177,
-      "learning_rate": 0.00039559999999999997,
-      "loss": 1.3997,
+      "grad_norm": 1.333957552909851,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 1.5851,
       "step": 1000
     },
     {
       "epoch": 1.6,
-      "eval_cer": 0.39621716426834613,
-      "eval_loss": 2.00065016746521,
-      "eval_runtime": 159.2575,
-      "eval_samples_per_second": 19.691,
-      "eval_steps_per_second": 1.231,
-      "eval_wer": 0.6068801160501502,
+      "eval_cer": 0.4133216406903974,
+      "eval_loss": 1.9947007894515991,
+      "eval_runtime": 158.5726,
+      "eval_samples_per_second": 19.776,
+      "eval_steps_per_second": 1.236,
+      "eval_wer": 0.6194798466480157,
       "step": 1000
     },
     {
       "epoch": 1.6016,
-      "grad_norm": 1.0269728899002075,
-      "learning_rate": 0.000396,
-      "loss": 1.5882,
+      "grad_norm": 0.975665807723999,
+      "learning_rate": 0.00029729999999999996,
+      "loss": 1.3631,
       "step": 1001
     },
     {
       "epoch": 1.6032,
-      "grad_norm": 1.0317885875701904,
-      "learning_rate": 0.00039639999999999993,
-      "loss": 1.7007,
+      "grad_norm": 1.125833511352539,
+      "learning_rate": 0.00029759999999999997,
+      "loss": 1.8541,
       "step": 1002
     },
     {
       "epoch": 1.6048,
-      "grad_norm": 1.605910062789917,
-      "learning_rate": 0.00039679999999999994,
-      "loss": 1.6842,
+      "grad_norm": 0.9652669429779053,
+      "learning_rate": 0.0002979,
+      "loss": 1.4736,
       "step": 1003
     },
     {
       "epoch": 1.6064,
-      "grad_norm": 1.5912996530532837,
-      "learning_rate": 0.0003972,
-      "loss": 1.3565,
+      "grad_norm": 1.7514461278915405,
+      "learning_rate": 0.0002982,
+      "loss": 1.3087,
       "step": 1004
     },
     {
       "epoch": 1.608,
-      "grad_norm": 0.761491596698761,
-      "learning_rate": 0.00039759999999999996,
-      "loss": 1.4429,
+      "grad_norm": 1.1740005016326904,
+      "learning_rate": 0.0002985,
+      "loss": 1.683,
       "step": 1005
     },
     {
       "epoch": 1.6096,
-      "grad_norm": 0.8185489177703857,
-      "learning_rate": 0.00039799999999999997,
-      "loss": 1.319,
+      "grad_norm": 1.2667077779769897,
+      "learning_rate": 0.0002988,
+      "loss": 1.3041,
       "step": 1006
     },
     {
       "epoch": 1.6112,
-      "grad_norm": 1.1462265253067017,
-      "learning_rate": 0.0003984,
-      "loss": 1.4704,
+      "grad_norm": 0.9430153965950012,
+      "learning_rate": 0.00029909999999999995,
+      "loss": 1.62,
       "step": 1007
     },
     {
       "epoch": 1.6128,
-      "grad_norm": 0.7877523899078369,
-      "learning_rate": 0.00039879999999999994,
-      "loss": 1.4737,
+      "grad_norm": 1.3095754384994507,
+      "learning_rate": 0.00029939999999999996,
+      "loss": 2.0639,
       "step": 1008
     },
     {
       "epoch": 1.6143999999999998,
-      "grad_norm": 1.2089191675186157,
-      "learning_rate": 0.00039919999999999995,
-      "loss": 1.8243,
+      "grad_norm": 1.265551209449768,
+      "learning_rate": 0.00029969999999999997,
+      "loss": 1.5279,
       "step": 1009
     },
     {
       "epoch": 1.616,
-      "grad_norm": 1.1152609586715698,
-      "learning_rate": 0.0003996,
-      "loss": 1.5541,
+      "grad_norm": 1.1508311033248901,
+      "learning_rate": 0.0003,
+      "loss": 1.6194,
       "step": 1010
     },
     {
       "epoch": 1.6176,
-      "grad_norm": 1.2637883424758911,
-      "learning_rate": 0.00039999999999999996,
-      "loss": 1.4111,
+      "grad_norm": 0.9043980836868286,
+      "learning_rate": 0.0003002999999999999,
+      "loss": 1.5519,
       "step": 1011
     },
     {
       "epoch": 1.6192,
-      "grad_norm": 1.272303819656372,
-      "learning_rate": 0.0004004,
-      "loss": 1.642,
+      "grad_norm": 1.0414581298828125,
+      "learning_rate": 0.0003006,
+      "loss": 1.4572,
       "step": 1012
     },
     {
       "epoch": 1.6208,
-      "grad_norm": 0.77174973487854,
-      "learning_rate": 0.0004008,
-      "loss": 1.6019,
+      "grad_norm": 1.113398790359497,
+      "learning_rate": 0.00030089999999999994,
+      "loss": 1.5507,
       "step": 1013
     },
     {
       "epoch": 1.6223999999999998,
-      "grad_norm": 1.0357544422149658,
-      "learning_rate": 0.00040119999999999994,
-      "loss": 1.7662,
+      "grad_norm": 1.0045056343078613,
+      "learning_rate": 0.00030119999999999995,
+      "loss": 1.6429,
       "step": 1014
     },
     {
       "epoch": 1.624,
-      "grad_norm": 1.2624030113220215,
-      "learning_rate": 0.00040159999999999995,
-      "loss": 1.44,
+      "grad_norm": 1.2001416683197021,
+      "learning_rate": 0.00030149999999999996,
+      "loss": 1.6436,
       "step": 1015
     },
     {
       "epoch": 1.6256,
-      "grad_norm": 2.911835193634033,
-      "learning_rate": 0.000402,
-      "loss": 1.9629,
+      "grad_norm": 1.195292353630066,
+      "learning_rate": 0.00030179999999999996,
+      "loss": 1.8505,
       "step": 1016
     },
     {
       "epoch": 1.6272,
-      "grad_norm": 1.212441325187683,
-      "learning_rate": 0.00040239999999999997,
-      "loss": 1.368,
+      "grad_norm": 1.6948364973068237,
+      "learning_rate": 0.0003020999999999999,
+      "loss": 1.4089,
       "step": 1017
     },
     {
       "epoch": 1.6288,
-      "grad_norm": 1.2743821144104004,
-      "learning_rate": 0.0004028,
-      "loss": 1.4819,
+      "grad_norm": 2.20973539352417,
+      "learning_rate": 0.0003024,
+      "loss": 2.0513,
       "step": 1018
     },
     {
       "epoch": 1.6303999999999998,
-      "grad_norm": 1.1178932189941406,
-      "learning_rate": 0.0004032,
-      "loss": 1.8235,
+      "grad_norm": 1.357258677482605,
+      "learning_rate": 0.00030269999999999993,
+      "loss": 1.5901,
       "step": 1019
     },
     {
       "epoch": 1.6320000000000001,
-      "grad_norm": 1.2212958335876465,
-      "learning_rate": 0.00040359999999999994,
-      "loss": 1.6722,
+      "grad_norm": 1.2716312408447266,
+      "learning_rate": 0.000303,
+      "loss": 1.8127,
       "step": 1020
     },
     {
       "epoch": 1.6336,
-      "grad_norm": 1.7435489892959595,
-      "learning_rate": 0.00040399999999999995,
-      "loss": 1.5568,
+      "grad_norm": 1.8802478313446045,
+      "learning_rate": 0.00030329999999999995,
+      "loss": 1.9268,
       "step": 1021
     },
     {
       "epoch": 1.6352,
-      "grad_norm": 1.944061040878296,
-      "learning_rate": 0.0004044,
-      "loss": 1.452,
+      "grad_norm": 1.2685855627059937,
+      "learning_rate": 0.00030359999999999995,
+      "loss": 1.8009,
       "step": 1022
     },
     {
       "epoch": 1.6368,
-      "grad_norm": 1.9267961978912354,
-      "learning_rate": 0.0004048,
-      "loss": 1.7871,
+      "grad_norm": 1.5773471593856812,
+      "learning_rate": 0.00030389999999999996,
+      "loss": 1.7688,
       "step": 1023
     },
     {
       "epoch": 1.6383999999999999,
-      "grad_norm": 2.1189751625061035,
-      "learning_rate": 0.0004052,
-      "loss": 2.1738,
+      "grad_norm": 1.9859917163848877,
+      "learning_rate": 0.00030419999999999997,
+      "loss": 2.0243,
       "step": 1024
     },
     {
       "epoch": 1.6400000000000001,
-      "grad_norm": 1.8044579029083252,
-      "learning_rate": 0.0004056,
-      "loss": 2.0679,
+      "grad_norm": 1.7131433486938477,
+      "learning_rate": 0.0003044999999999999,
+      "loss": 1.8381,
       "step": 1025
     },
     {
       "epoch": 1.6416,
-      "grad_norm": 9.363924980163574,
-      "learning_rate": 0.00040599999999999995,
-      "loss": 2.8019,
+      "grad_norm": 10.376376152038574,
+      "learning_rate": 0.0003048,
+      "loss": 3.4879,
       "step": 1026
     },
     {
       "epoch": 1.6432,
-      "grad_norm": 4.761209964752197,
-      "learning_rate": 0.00040639999999999996,
-      "loss": 2.2336,
+      "grad_norm": 5.2158284187316895,
+      "learning_rate": 0.00030509999999999994,
+      "loss": 2.5737,
       "step": 1027
     },
     {
       "epoch": 1.6448,
-      "grad_norm": 7.563479900360107,
-      "learning_rate": 0.00040679999999999997,
-      "loss": 2.7444,
+      "grad_norm": 7.62057638168335,
+      "learning_rate": 0.0003054,
+      "loss": 3.0633,
       "step": 1028
     },
     {
       "epoch": 1.6463999999999999,
-      "grad_norm": 3.261016845703125,
-      "learning_rate": 0.0004071999999999999,
-      "loss": 1.9803,
+      "grad_norm": 2.5151166915893555,
+      "learning_rate": 0.00030569999999999995,
+      "loss": 1.9631,
       "step": 1029
     },
     {
       "epoch": 1.6480000000000001,
-      "grad_norm": 2.0877084732055664,
-      "learning_rate": 0.0004076,
-      "loss": 1.787,
+      "grad_norm": 2.6367363929748535,
+      "learning_rate": 0.00030599999999999996,
+      "loss": 2.0063,
       "step": 1030
     },
     {
       "epoch": 1.6496,
-      "grad_norm": 1.2679075002670288,
-      "learning_rate": 0.000408,
-      "loss": 1.6983,
+      "grad_norm": 2.630662679672241,
+      "learning_rate": 0.00030629999999999996,
+      "loss": 1.661,
       "step": 1031
     },
     {
       "epoch": 1.6512,
-      "grad_norm": 2.241767168045044,
-      "learning_rate": 0.00040839999999999995,
-      "loss": 1.8019,
+      "grad_norm": 3.1744046211242676,
+      "learning_rate": 0.00030659999999999997,
+      "loss": 1.9693,
       "step": 1032
     },
     {
       "epoch": 1.6528,
-      "grad_norm": 2.6402766704559326,
-      "learning_rate": 0.00040879999999999996,
-      "loss": 1.6751,
+      "grad_norm": 3.2799153327941895,
+      "learning_rate": 0.0003068999999999999,
+      "loss": 1.8917,
       "step": 1033
     },
     {
       "epoch": 1.6543999999999999,
-      "grad_norm": 0.9462832808494568,
-      "learning_rate": 0.00040919999999999997,
-      "loss": 1.9119,
+      "grad_norm": 2.914391279220581,
+      "learning_rate": 0.0003072,
+      "loss": 1.8649,
       "step": 1034
     },
     {
       "epoch": 1.6560000000000001,
-      "grad_norm": 1.57070791721344,
-      "learning_rate": 0.00040959999999999993,
-      "loss": 1.4415,
+      "grad_norm": 1.2659504413604736,
+      "learning_rate": 0.00030749999999999994,
+      "loss": 1.7872,
       "step": 1035
     },
     {
       "epoch": 1.6576,
-      "grad_norm": 0.9740553498268127,
-      "learning_rate": 0.00041,
-      "loss": 1.5275,
+      "grad_norm": 1.2379803657531738,
+      "learning_rate": 0.0003078,
+      "loss": 1.4999,
       "step": 1036
     },
     {
       "epoch": 1.6592,
-      "grad_norm": 2.7589497566223145,
-      "learning_rate": 0.0004104,
-      "loss": 1.6575,
+      "grad_norm": 2.3075740337371826,
+      "learning_rate": 0.00030809999999999995,
+      "loss": 1.6165,
       "step": 1037
     },
     {
       "epoch": 1.6608,
-      "grad_norm": 2.4938724040985107,
-      "learning_rate": 0.00041079999999999996,
-      "loss": 1.8596,
+      "grad_norm": 2.4933927059173584,
+      "learning_rate": 0.00030839999999999996,
+      "loss": 1.5173,
       "step": 1038
     },
     {
       "epoch": 1.6623999999999999,
-      "grad_norm": 1.9689472913742065,
-      "learning_rate": 0.00041119999999999997,
-      "loss": 1.3647,
+      "grad_norm": 1.777960181236267,
+      "learning_rate": 0.00030869999999999997,
+      "loss": 1.3944,
       "step": 1039
     },
     {
       "epoch": 1.6640000000000001,
-      "grad_norm": 2.6272342205047607,
-      "learning_rate": 0.0004116,
-      "loss": 1.4989,
+      "grad_norm": 2.229851484298706,
+      "learning_rate": 0.000309,
+      "loss": 1.6344,
       "step": 1040
     },
     {
       "epoch": 1.6656,
-      "grad_norm": 1.4329277276992798,
-      "learning_rate": 0.00041199999999999993,
-      "loss": 1.2195,
+      "grad_norm": 1.5905592441558838,
+      "learning_rate": 0.00030929999999999993,
+      "loss": 1.4025,
       "step": 1041
     },
     {
       "epoch": 1.6672,
-      "grad_norm": 1.3667677640914917,
-      "learning_rate": 0.0004124,
-      "loss": 1.3946,
+      "grad_norm": 0.9465227723121643,
+      "learning_rate": 0.0003096,
+      "loss": 1.5164,
       "step": 1042
     },
     {
       "epoch": 1.6688,
-      "grad_norm": 0.6925958395004272,
-      "learning_rate": 0.00041279999999999995,
-      "loss": 1.3634,
+      "grad_norm": 1.2524248361587524,
+      "learning_rate": 0.00030989999999999994,
+      "loss": 1.5344,
       "step": 1043
     },
     {
       "epoch": 1.6703999999999999,
-      "grad_norm": 0.9826893210411072,
-      "learning_rate": 0.00041319999999999996,
-      "loss": 1.4918,
+      "grad_norm": 0.9570756554603577,
+      "learning_rate": 0.0003102,
+      "loss": 1.4809,
       "step": 1044
     },
     {
       "epoch": 1.6720000000000002,
-      "grad_norm": 1.4587767124176025,
-      "learning_rate": 0.00041359999999999997,
-      "loss": 1.6036,
+      "grad_norm": 1.3065333366394043,
+      "learning_rate": 0.00031049999999999996,
+      "loss": 1.3097,
       "step": 1045
     },
     {
       "epoch": 1.6736,
-      "grad_norm": 0.8268725872039795,
-      "learning_rate": 0.0004139999999999999,
-      "loss": 1.5278,
+      "grad_norm": 0.7914066314697266,
+      "learning_rate": 0.00031079999999999997,
+      "loss": 1.5068,
       "step": 1046
     },
     {
       "epoch": 1.6752,
-      "grad_norm": 0.8046411871910095,
-      "learning_rate": 0.00041439999999999994,
-      "loss": 1.1556,
+      "grad_norm": 1.5971769094467163,
+      "learning_rate": 0.00031109999999999997,
+      "loss": 1.7266,
       "step": 1047
     },
     {
       "epoch": 1.6768,
-      "grad_norm": 0.6971915364265442,
-      "learning_rate": 0.0004148,
-      "loss": 1.6188,
+      "grad_norm": 2.6839516162872314,
+      "learning_rate": 0.0003114,
+      "loss": 1.7296,
       "step": 1048
     },
     {
       "epoch": 1.6784,
-      "grad_norm": 2.461832284927368,
-      "learning_rate": 0.00041519999999999995,
-      "loss": 1.4803,
+      "grad_norm": 1.134955883026123,
+      "learning_rate": 0.00031169999999999993,
+      "loss": 1.1195,
       "step": 1049
     },
     {
       "epoch": 1.6800000000000002,
-      "grad_norm": 1.7602179050445557,
-      "learning_rate": 0.00041559999999999996,
-      "loss": 1.4928,
+      "grad_norm": 0.9906765222549438,
+      "learning_rate": 0.000312,
+      "loss": 1.4243,
       "step": 1050
     },
     {
       "epoch": 1.6816,
-      "grad_norm": 0.8309889435768127,
-      "learning_rate": 0.000416,
-      "loss": 1.4609,
+      "grad_norm": 1.2784078121185303,
+      "learning_rate": 0.00031229999999999995,
+      "loss": 1.5288,
       "step": 1051
     },
     {
       "epoch": 1.6832,
-      "grad_norm": 1.2826169729232788,
-      "learning_rate": 0.00041639999999999993,
-      "loss": 1.5435,
+      "grad_norm": 0.8589702844619751,
+      "learning_rate": 0.0003126,
+      "loss": 1.5027,
       "step": 1052
     },
     {
       "epoch": 1.6848,
-      "grad_norm": 0.8392765522003174,
-      "learning_rate": 0.00041679999999999994,
-      "loss": 1.326,
+      "grad_norm": 0.9791033864021301,
+      "learning_rate": 0.00031289999999999996,
+      "loss": 1.3079,
       "step": 1053
     },
     {
       "epoch": 1.6864,
-      "grad_norm": 0.830362856388092,
-      "learning_rate": 0.0004172,
-      "loss": 1.5522,
+      "grad_norm": 1.6374212503433228,
+      "learning_rate": 0.00031319999999999997,
+      "loss": 1.7103,
       "step": 1054
     },
     {
       "epoch": 1.688,
-      "grad_norm": 1.0867887735366821,
-      "learning_rate": 0.00041759999999999996,
-      "loss": 1.5438,
+      "grad_norm": 0.799659252166748,
+      "learning_rate": 0.0003135,
+      "loss": 1.6644,
       "step": 1055
     },
     {
       "epoch": 1.6896,
-      "grad_norm": 1.0517538785934448,
-      "learning_rate": 0.00041799999999999997,
-      "loss": 1.3964,
+      "grad_norm": 1.1091631650924683,
+      "learning_rate": 0.0003138,
+      "loss": 1.3437,
       "step": 1056
     },
     {
       "epoch": 1.6912,
-      "grad_norm": 1.4779033660888672,
-      "learning_rate": 0.0004184,
-      "loss": 1.5742,
+      "grad_norm": 1.3691285848617554,
+      "learning_rate": 0.00031409999999999994,
+      "loss": 1.3581,
       "step": 1057
     },
     {
       "epoch": 1.6928,
-      "grad_norm": 0.9230872392654419,
-      "learning_rate": 0.00041879999999999993,
-      "loss": 1.3136,
+      "grad_norm": 1.286253571510315,
+      "learning_rate": 0.0003144,
+      "loss": 1.3635,
       "step": 1058
     },
     {
       "epoch": 1.6944,
-      "grad_norm": 1.0470373630523682,
-      "learning_rate": 0.00041919999999999994,
-      "loss": 1.7794,
+      "grad_norm": 0.8725206851959229,
+      "learning_rate": 0.00031469999999999995,
+      "loss": 1.7202,
       "step": 1059
     },
     {
       "epoch": 1.696,
-      "grad_norm": 1.1661121845245361,
-      "learning_rate": 0.0004196,
-      "loss": 1.6266,
+      "grad_norm": 1.2710314989089966,
+      "learning_rate": 0.00031499999999999996,
+      "loss": 1.4148,
       "step": 1060
     },
     {
       "epoch": 1.6976,
-      "grad_norm": 0.8189252614974976,
-      "learning_rate": 0.00041999999999999996,
-      "loss": 1.2236,
+      "grad_norm": 1.11667799949646,
+      "learning_rate": 0.00031529999999999997,
+      "loss": 1.226,
       "step": 1061
     },
     {
       "epoch": 1.6992,
-      "grad_norm": 1.0342382192611694,
-      "learning_rate": 0.0004204,
-      "loss": 1.4281,
+      "grad_norm": 1.9654427766799927,
+      "learning_rate": 0.0003156,
+      "loss": 1.8421,
       "step": 1062
     },
     {
       "epoch": 1.7008,
-      "grad_norm": 0.9032308459281921,
-      "learning_rate": 0.0004208,
-      "loss": 1.5472,
+      "grad_norm": 1.2792798280715942,
+      "learning_rate": 0.0003158999999999999,
+      "loss": 1.6172,
       "step": 1063
     },
     {
       "epoch": 1.7024,
-      "grad_norm": 0.8276682496070862,
-      "learning_rate": 0.00042119999999999994,
-      "loss": 1.397,
+      "grad_norm": 1.032118558883667,
+      "learning_rate": 0.0003162,
+      "loss": 1.5739,
       "step": 1064
     },
     {
       "epoch": 1.704,
-      "grad_norm": 1.0572770833969116,
-      "learning_rate": 0.00042159999999999995,
-      "loss": 1.6688,
+      "grad_norm": 1.32432222366333,
+      "learning_rate": 0.00031649999999999994,
+      "loss": 1.6586,
       "step": 1065
     },
     {
       "epoch": 1.7056,
-      "grad_norm": 1.0404248237609863,
-      "learning_rate": 0.00042199999999999996,
-      "loss": 1.5673,
+      "grad_norm": 1.1895533800125122,
+      "learning_rate": 0.0003168,
+      "loss": 1.6227,
       "step": 1066
     },
     {
       "epoch": 1.7072,
-      "grad_norm": 1.2805649042129517,
-      "learning_rate": 0.0004223999999999999,
-      "loss": 1.6057,
+      "grad_norm": 1.4331198930740356,
+      "learning_rate": 0.00031709999999999996,
+      "loss": 1.5119,
       "step": 1067
     },
     {
       "epoch": 1.7088,
-      "grad_norm": 1.1005247831344604,
-      "learning_rate": 0.0004228,
-      "loss": 1.0563,
+      "grad_norm": 1.058091163635254,
+      "learning_rate": 0.00031739999999999996,
+      "loss": 1.7509,
       "step": 1068
     },
     {
       "epoch": 1.7104,
-      "grad_norm": 1.6212459802627563,
-      "learning_rate": 0.0004232,
-      "loss": 1.4625,
+      "grad_norm": 0.9952759742736816,
+      "learning_rate": 0.00031769999999999997,
+      "loss": 2.0645,
       "step": 1069
     },
     {
       "epoch": 1.712,
-      "grad_norm": 1.2400896549224854,
-      "learning_rate": 0.00042359999999999994,
-      "loss": 1.9359,
+      "grad_norm": 1.0876610279083252,
+      "learning_rate": 0.000318,
+      "loss": 1.5768,
       "step": 1070
     },
     {
       "epoch": 1.7136,
-      "grad_norm": 1.3954520225524902,
-      "learning_rate": 0.00042399999999999995,
-      "loss": 1.6781,
+      "grad_norm": 1.2695331573486328,
+      "learning_rate": 0.00031829999999999993,
+      "loss": 1.549,
       "step": 1071
     },
     {
       "epoch": 1.7151999999999998,
-      "grad_norm": 1.5502781867980957,
-      "learning_rate": 0.00042439999999999996,
-      "loss": 1.6238,
+      "grad_norm": 1.5271435976028442,
+      "learning_rate": 0.0003186,
+      "loss": 1.9173,
       "step": 1072
     },
     {
       "epoch": 1.7168,
-      "grad_norm": 1.2833088636398315,
-      "learning_rate": 0.0004247999999999999,
-      "loss": 1.9013,
+      "grad_norm": 1.755866527557373,
+      "learning_rate": 0.00031889999999999995,
+      "loss": 2.0375,
       "step": 1073
     },
     {
       "epoch": 1.7184,
-      "grad_norm": 1.4674335718154907,
-      "learning_rate": 0.0004252,
-      "loss": 1.531,
+      "grad_norm": 1.5030609369277954,
+      "learning_rate": 0.0003192,
+      "loss": 2.4663,
       "step": 1074
     },
     {
       "epoch": 1.72,
       "grad_norm": NaN,
-      "learning_rate": 0.0004252,
-      "loss": 1.2603,
+      "learning_rate": 0.0003192,
+      "loss": 1.5195,
       "step": 1075
     },
     {
       "epoch": 1.7216,
-      "grad_norm": 9.69878101348877,
-      "learning_rate": 0.0004256,
-      "loss": 3.4215,
+      "grad_norm": 6.813086032867432,
+      "learning_rate": 0.00031949999999999996,
+      "loss": 2.6142,
       "step": 1076
     },
     {
       "epoch": 1.7231999999999998,
-      "grad_norm": 4.316009998321533,
-      "learning_rate": 0.00042599999999999995,
-      "loss": 2.3031,
+      "grad_norm": 6.0582990646362305,
+      "learning_rate": 0.00031979999999999997,
+      "loss": 3.1514,
       "step": 1077
     },
     {
       "epoch": 1.7248,
-      "grad_norm": 2.6098554134368896,
-      "learning_rate": 0.00042639999999999996,
-      "loss": 2.5445,
+      "grad_norm": 1.800521969795227,
+      "learning_rate": 0.0003201,
+      "loss": 2.0795,
       "step": 1078
     },
     {
       "epoch": 1.7264,
-      "grad_norm": 2.034123182296753,
-      "learning_rate": 0.00042679999999999997,
-      "loss": 1.8094,
+      "grad_norm": 1.409582495689392,
+      "learning_rate": 0.0003204,
+      "loss": 1.9426,
       "step": 1079
     },
     {
       "epoch": 1.728,
-      "grad_norm": 2.0914783477783203,
-      "learning_rate": 0.0004271999999999999,
-      "loss": 1.9322,
+      "grad_norm": 2.2173216342926025,
+      "learning_rate": 0.00032069999999999993,
+      "loss": 1.8343,
       "step": 1080
     },
     {
       "epoch": 1.7296,
-      "grad_norm": 2.9455251693725586,
-      "learning_rate": 0.0004276,
-      "loss": 2.1734,
+      "grad_norm": 1.2972121238708496,
+      "learning_rate": 0.000321,
+      "loss": 2.2537,
       "step": 1081
     },
     {
       "epoch": 1.7311999999999999,
-      "grad_norm": 2.6501622200012207,
-      "learning_rate": 0.000428,
-      "loss": 2.0604,
+      "grad_norm": 2.211015224456787,
+      "learning_rate": 0.00032129999999999995,
+      "loss": 1.8755,
       "step": 1082
     },
     {
       "epoch": 1.7328000000000001,
-      "grad_norm": 1.1894841194152832,
-      "learning_rate": 0.00042839999999999995,
-      "loss": 2.3638,
+      "grad_norm": 1.617491364479065,
+      "learning_rate": 0.0003216,
+      "loss": 1.6271,
       "step": 1083
     },
     {
       "epoch": 1.7344,
-      "grad_norm": 1.127158284187317,
-      "learning_rate": 0.00042879999999999996,
-      "loss": 1.8291,
+      "grad_norm": 1.3888314962387085,
+      "learning_rate": 0.00032189999999999996,
+      "loss": 1.6329,
       "step": 1084
     },
     {
       "epoch": 1.736,
-      "grad_norm": 1.0093775987625122,
-      "learning_rate": 0.00042919999999999997,
-      "loss": 1.6005,
+      "grad_norm": 2.2377328872680664,
+      "learning_rate": 0.00032219999999999997,
+      "loss": 1.8106,
       "step": 1085
     },
     {
       "epoch": 1.7376,
-      "grad_norm": 1.7011162042617798,
-      "learning_rate": 0.0004295999999999999,
-      "loss": 1.5392,
+      "grad_norm": 1.4609044790267944,
+      "learning_rate": 0.0003225,
+      "loss": 1.2795,
       "step": 1086
     },
     {
       "epoch": 1.7391999999999999,
-      "grad_norm": 5.270832538604736,
-      "learning_rate": 0.00043,
-      "loss": 2.0137,
+      "grad_norm": 1.2239532470703125,
+      "learning_rate": 0.0003228,
+      "loss": 1.7129,
       "step": 1087
     },
     {
       "epoch": 1.7408000000000001,
-      "grad_norm": 4.517563343048096,
-      "learning_rate": 0.0004304,
-      "loss": 1.8987,
+      "grad_norm": 1.6309138536453247,
+      "learning_rate": 0.00032309999999999994,
+      "loss": 1.3214,
       "step": 1088
     },
     {
       "epoch": 1.7424,
-      "grad_norm": 2.386322498321533,
-      "learning_rate": 0.00043079999999999995,
-      "loss": 1.659,
+      "grad_norm": 2.0995709896087646,
+      "learning_rate": 0.0003234,
+      "loss": 1.5447,
       "step": 1089
     },
     {
       "epoch": 1.744,
-      "grad_norm": 2.8278114795684814,
-      "learning_rate": 0.00043119999999999996,
-      "loss": 1.7082,
+      "grad_norm": 1.7949155569076538,
+      "learning_rate": 0.00032369999999999995,
+      "loss": 1.4792,
       "step": 1090
     },
     {
       "epoch": 1.7456,
-      "grad_norm": 2.6224570274353027,
-      "learning_rate": 0.0004316,
-      "loss": 1.844,
+      "grad_norm": 0.9978499412536621,
+      "learning_rate": 0.000324,
+      "loss": 1.2458,
       "step": 1091
     },
     {
       "epoch": 1.7471999999999999,
-      "grad_norm": 0.962620198726654,
-      "learning_rate": 0.00043199999999999993,
-      "loss": 1.74,
+      "grad_norm": 0.9504018425941467,
+      "learning_rate": 0.00032429999999999997,
+      "loss": 1.4711,
       "step": 1092
     },
     {
       "epoch": 1.7488000000000001,
-      "grad_norm": 1.115366816520691,
-      "learning_rate": 0.0004324,
-      "loss": 1.781,
+      "grad_norm": 1.0632179975509644,
+      "learning_rate": 0.0003246,
+      "loss": 1.5718,
       "step": 1093
     },
     {
       "epoch": 1.7504,
-      "grad_norm": 1.3309029340744019,
-      "learning_rate": 0.0004328,
-      "loss": 1.7973,
+      "grad_norm": 1.2612885236740112,
+      "learning_rate": 0.0003249,
+      "loss": 1.542,
       "step": 1094
     },
     {
       "epoch": 1.752,
-      "grad_norm": 1.3224897384643555,
-      "learning_rate": 0.00043319999999999996,
-      "loss": 1.6193,
+      "grad_norm": 1.4393160343170166,
+      "learning_rate": 0.0003252,
+      "loss": 1.4533,
       "step": 1095
     },
     {
       "epoch": 1.7536,
-      "grad_norm": 0.8370805382728577,
-      "learning_rate": 0.00043359999999999997,
-      "loss": 1.541,
+      "grad_norm": 1.2894624471664429,
+      "learning_rate": 0.00032549999999999994,
+      "loss": 1.2829,
       "step": 1096
     },
     {
       "epoch": 1.7551999999999999,
-      "grad_norm": 1.5298939943313599,
-      "learning_rate": 0.000434,
-      "loss": 1.5882,
+      "grad_norm": 0.9237915873527527,
+      "learning_rate": 0.0003258,
+      "loss": 1.5936,
       "step": 1097
     },
     {
       "epoch": 1.7568000000000001,
-      "grad_norm": 2.6840977668762207,
-      "learning_rate": 0.00043439999999999993,
-      "loss": 1.9575,
+      "grad_norm": 0.7497216463088989,
+      "learning_rate": 0.00032609999999999996,
+      "loss": 1.6193,
       "step": 1098
     },
     {
       "epoch": 1.7584,
-      "grad_norm": 1.4588136672973633,
-      "learning_rate": 0.0004348,
-      "loss": 2.0164,
+      "grad_norm": 2.3699493408203125,
+      "learning_rate": 0.0003264,
+      "loss": 1.4169,
       "step": 1099
     },
     {
       "epoch": 1.76,
-      "grad_norm": 1.8328728675842285,
-      "learning_rate": 0.0004352,
-      "loss": 1.768,
-      "step": 1100
-    },
-    {
-      "epoch": 1.76,
-      "eval_cer": 0.41231713636544826,
-      "eval_loss": 2.0711567401885986,
-      "eval_runtime": 157.9834,
-      "eval_samples_per_second": 19.85,
-      "eval_steps_per_second": 1.241,
-      "eval_wer": 0.6448243705315512,
+      "grad_norm": 1.4830172061920166,
+      "learning_rate": 0.00032669999999999997,
+      "loss": 1.3362,
       "step": 1100
     },
     {
       "epoch": 1.7616,
-      "grad_norm": 1.0604596138000488,
-      "learning_rate": 0.00043559999999999996,
-      "loss": 1.4263,
+      "grad_norm": 1.660394310951233,
+      "learning_rate": 0.000327,
+      "loss": 1.5181,
       "step": 1101
     },
     {
       "epoch": 1.7631999999999999,
-      "grad_norm": 0.9676076769828796,
-      "learning_rate": 0.00043599999999999997,
-      "loss": 1.6473,
+      "grad_norm": 1.3307558298110962,
+      "learning_rate": 0.0003273,
+      "loss": 1.5403,
       "step": 1102
     },
     {
       "epoch": 1.7648000000000001,
-      "grad_norm": 0.9821343421936035,
-      "learning_rate": 0.0004364,
-      "loss": 1.8823,
+      "grad_norm": 1.7247034311294556,
+      "learning_rate": 0.0003276,
+      "loss": 1.5718,
       "step": 1103
     },
     {
       "epoch": 1.7664,
-      "grad_norm": 0.9323663711547852,
-      "learning_rate": 0.00043679999999999994,
-      "loss": 1.5884,
+      "grad_norm": 1.1111986637115479,
+      "learning_rate": 0.00032789999999999995,
+      "loss": 1.3949,
       "step": 1104
     },
     {
       "epoch": 1.768,
-      "grad_norm": 0.6213338971138,
-      "learning_rate": 0.00043719999999999995,
-      "loss": 1.5625,
+      "grad_norm": 1.0019989013671875,
+      "learning_rate": 0.0003282,
+      "loss": 1.486,
       "step": 1105
     },
     {
       "epoch": 1.7696,
-      "grad_norm": 0.767331063747406,
-      "learning_rate": 0.0004375999999999999,
-      "loss": 1.5178,
+      "grad_norm": 0.9997567534446716,
+      "learning_rate": 0.00032849999999999996,
+      "loss": 1.4654,
       "step": 1106
     },
     {
       "epoch": 1.7711999999999999,
-      "grad_norm": 1.7353647947311401,
-      "learning_rate": 0.00043799999999999997,
-      "loss": 1.7371,
+      "grad_norm": 0.9817771315574646,
+      "learning_rate": 0.0003288,
+      "loss": 1.4524,
       "step": 1107
     },
     {
       "epoch": 1.7728000000000002,
-      "grad_norm": 1.3577700853347778,
-      "learning_rate": 0.0004384,
-      "loss": 1.5416,
+      "grad_norm": 1.1444945335388184,
+      "learning_rate": 0.0003291,
+      "loss": 1.525,
       "step": 1108
     },
     {
       "epoch": 1.7744,
-      "grad_norm": 1.7404487133026123,
-      "learning_rate": 0.00043879999999999993,
-      "loss": 1.4647,
+      "grad_norm": 1.1713188886642456,
+      "learning_rate": 0.0003294,
+      "loss": 1.4727,
       "step": 1109
     },
     {
       "epoch": 1.776,
-      "grad_norm": 2.192164659500122,
-      "learning_rate": 0.00043919999999999994,
-      "loss": 1.7362,
+      "grad_norm": 1.6740111112594604,
+      "learning_rate": 0.0003297,
+      "loss": 1.6013,
       "step": 1110
     },
     {
       "epoch": 1.7776,
-      "grad_norm": 0.7440006732940674,
-      "learning_rate": 0.00043959999999999995,
-      "loss": 1.8019,
+      "grad_norm": 1.3131376504898071,
+      "learning_rate": 0.00033,
+      "loss": 1.5525,
       "step": 1111
     },
     {
       "epoch": 1.7792,
-      "grad_norm": 0.8111011385917664,
-      "learning_rate": 0.0004399999999999999,
-      "loss": 1.6845,
+      "grad_norm": 1.5951952934265137,
+      "learning_rate": 0.00033029999999999995,
+      "loss": 1.5049,
       "step": 1112
     },
     {
       "epoch": 1.7808000000000002,
-      "grad_norm": 0.8545581102371216,
-      "learning_rate": 0.00044039999999999997,
-      "loss": 1.8391,
+      "grad_norm": 1.0123345851898193,
+      "learning_rate": 0.0003306,
+      "loss": 1.655,
       "step": 1113
     },
     {
       "epoch": 1.7824,
-      "grad_norm": 1.5863672494888306,
-      "learning_rate": 0.0004408,
-      "loss": 1.8954,
+      "grad_norm": 0.9980636239051819,
+      "learning_rate": 0.00033089999999999997,
+      "loss": 1.413,
       "step": 1114
     },
     {
       "epoch": 1.784,
-      "grad_norm": 0.9695616960525513,
-      "learning_rate": 0.00044119999999999994,
-      "loss": 1.4524,
+      "grad_norm": 1.0149414539337158,
+      "learning_rate": 0.0003312,
+      "loss": 1.3944,
       "step": 1115
     },
     {
       "epoch": 1.7856,
-      "grad_norm": 0.8583353757858276,
-      "learning_rate": 0.00044159999999999995,
-      "loss": 1.91,
+      "grad_norm": 1.0100555419921875,
+      "learning_rate": 0.0003315,
+      "loss": 1.6389,
       "step": 1116
     },
     {
       "epoch": 1.7872,
-      "grad_norm": 1.3024649620056152,
-      "learning_rate": 0.00044199999999999996,
-      "loss": 2.0295,
+      "grad_norm": 1.0911948680877686,
+      "learning_rate": 0.0003318,
+      "loss": 1.6691,
       "step": 1117
     },
     {
       "epoch": 1.7888,
-      "grad_norm": 1.960022211074829,
-      "learning_rate": 0.0004423999999999999,
-      "loss": 1.8059,
+      "grad_norm": 0.9347399473190308,
+      "learning_rate": 0.00033209999999999994,
+      "loss": 2.1185,
       "step": 1118
     },
     {
       "epoch": 1.7904,
-      "grad_norm": 0.7005279660224915,
-      "learning_rate": 0.0004428,
-      "loss": 1.7367,
+      "grad_norm": 0.9940705895423889,
+      "learning_rate": 0.0003324,
+      "loss": 1.8248,
       "step": 1119
     },
     {
       "epoch": 1.792,
-      "grad_norm": 1.53810453414917,
-      "learning_rate": 0.0004432,
-      "loss": 1.683,
+      "grad_norm": 1.3178499937057495,
+      "learning_rate": 0.00033269999999999996,
+      "loss": 1.3904,
       "step": 1120
     },
     {
       "epoch": 1.7936,
-      "grad_norm": 1.2621592283248901,
-      "learning_rate": 0.00044359999999999994,
-      "loss": 1.8247,
+      "grad_norm": 1.338726282119751,
+      "learning_rate": 0.000333,
+      "loss": 1.6328,
       "step": 1121
     },
     {
       "epoch": 1.7952,
-      "grad_norm": 1.0741353034973145,
-      "learning_rate": 0.00044399999999999995,
-      "loss": 1.6832,
+      "grad_norm": 1.6837692260742188,
+      "learning_rate": 0.00033329999999999997,
+      "loss": 1.8429,
       "step": 1122
     },
     {
       "epoch": 1.7968,
-      "grad_norm": 1.36100172996521,
-      "learning_rate": 0.00044439999999999996,
-      "loss": 1.4735,
+      "grad_norm": 1.9074715375900269,
+      "learning_rate": 0.0003336,
+      "loss": 1.9857,
       "step": 1123
     },
     {
       "epoch": 1.7984,
-      "grad_norm": 1.0110926628112793,
-      "learning_rate": 0.0004447999999999999,
-      "loss": 1.9363,
+      "grad_norm": 1.2642357349395752,
+      "learning_rate": 0.0003339,
+      "loss": 2.0672,
       "step": 1124
     },
     {
       "epoch": 1.8,
-      "grad_norm": 1.4565409421920776,
-      "learning_rate": 0.0004452,
-      "loss": 2.7498,
+      "grad_norm": 1.70827054977417,
+      "learning_rate": 0.0003342,
+      "loss": 2.2459,
       "step": 1125
     },
     {
       "epoch": 1.8016,
-      "grad_norm": 13.088458061218262,
-      "learning_rate": 0.0004456,
-      "loss": 3.6198,
+      "grad_norm": 8.30029582977295,
+      "learning_rate": 0.00033449999999999994,
+      "loss": 3.0681,
       "step": 1126
     },
     {
       "epoch": 1.8032,
-      "grad_norm": 2.855563163757324,
-      "learning_rate": 0.00044599999999999994,
-      "loss": 2.1456,
+      "grad_norm": 3.6573855876922607,
+      "learning_rate": 0.0003348,
+      "loss": 2.4,
       "step": 1127
     },
     {
       "epoch": 1.8048,
-      "grad_norm": 1.2719941139221191,
-      "learning_rate": 0.00044639999999999995,
-      "loss": 1.972,
+      "grad_norm": 1.3085684776306152,
+      "learning_rate": 0.00033509999999999996,
+      "loss": 2.0479,
       "step": 1128
     },
     {
       "epoch": 1.8064,
-      "grad_norm": 2.133577585220337,
-      "learning_rate": 0.00044679999999999996,
-      "loss": 1.8491,
+      "grad_norm": 1.3093235492706299,
+      "learning_rate": 0.0003354,
+      "loss": 1.6513,
       "step": 1129
     },
     {
       "epoch": 1.808,
-      "grad_norm": 2.3331825733184814,
-      "learning_rate": 0.0004471999999999999,
-      "loss": 2.1332,
+      "grad_norm": 2.186389207839966,
+      "learning_rate": 0.0003357,
+      "loss": 1.64,
       "step": 1130
     },
     {
       "epoch": 1.8096,
-      "grad_norm": 2.8282387256622314,
-      "learning_rate": 0.0004476,
-      "loss": 2.1503,
+      "grad_norm": 2.4307668209075928,
+      "learning_rate": 0.000336,
+      "loss": 1.7372,
       "step": 1131
     },
     {
       "epoch": 1.8112,
-      "grad_norm": 1.7458895444869995,
-      "learning_rate": 0.000448,
-      "loss": 2.0523,
+      "grad_norm": 1.5897157192230225,
+      "learning_rate": 0.0003363,
+      "loss": 1.6175,
       "step": 1132
     },
     {
       "epoch": 1.8128,
-      "grad_norm": 1.0502371788024902,
-      "learning_rate": 0.00044839999999999995,
-      "loss": 1.8634,
+      "grad_norm": 1.1186329126358032,
+      "learning_rate": 0.0003366,
+      "loss": 1.5705,
       "step": 1133
     },
     {
       "epoch": 1.8144,
-      "grad_norm": 1.3230615854263306,
-      "learning_rate": 0.00044879999999999996,
-      "loss": 2.218,
+      "grad_norm": 1.862901210784912,
+      "learning_rate": 0.00033689999999999995,
+      "loss": 1.4379,
       "step": 1134
     },
     {
       "epoch": 1.8159999999999998,
-      "grad_norm": 2.9722142219543457,
-      "learning_rate": 0.00044919999999999997,
-      "loss": 1.9498,
+      "grad_norm": 3.026568651199341,
+      "learning_rate": 0.0003372,
+      "loss": 1.7887,
       "step": 1135
     },
     {
       "epoch": 1.8176,
-      "grad_norm": 5.391232967376709,
-      "learning_rate": 0.0004495999999999999,
-      "loss": 2.1481,
+      "grad_norm": 3.224273204803467,
+      "learning_rate": 0.00033749999999999996,
+      "loss": 1.5298,
       "step": 1136
     },
     {
       "epoch": 1.8192,
-      "grad_norm": 8.026174545288086,
-      "learning_rate": 0.00045,
-      "loss": 2.5871,
+      "grad_norm": 1.6268669366836548,
+      "learning_rate": 0.0003377999999999999,
+      "loss": 1.2549,
       "step": 1137
     },
     {
       "epoch": 1.8208,
-      "grad_norm": 5.394509315490723,
-      "learning_rate": 0.0004504,
-      "loss": 2.076,
+      "grad_norm": 1.8013404607772827,
+      "learning_rate": 0.0003381,
+      "loss": 1.3867,
       "step": 1138
     },
     {
       "epoch": 1.8224,
-      "grad_norm": 4.2326579093933105,
-      "learning_rate": 0.00045079999999999995,
-      "loss": 2.1723,
+      "grad_norm": 1.5397872924804688,
+      "learning_rate": 0.00033839999999999993,
+      "loss": 1.7614,
       "step": 1139
     },
     {
       "epoch": 1.8239999999999998,
-      "grad_norm": 2.653384208679199,
-      "learning_rate": 0.00045119999999999996,
-      "loss": 1.9721,
+      "grad_norm": 0.8708004355430603,
+      "learning_rate": 0.0003387,
+      "loss": 1.1885,
       "step": 1140
     },
     {
       "epoch": 1.8256000000000001,
-      "grad_norm": 2.5420026779174805,
-      "learning_rate": 0.00045159999999999997,
-      "loss": 1.8613,
+      "grad_norm": 0.7449163198471069,
+      "learning_rate": 0.00033899999999999995,
+      "loss": 1.4271,
       "step": 1141
     },
     {
       "epoch": 1.8272,
-      "grad_norm": 1.363802433013916,
-      "learning_rate": 0.00045199999999999993,
-      "loss": 2.0218,
+      "grad_norm": 0.8833471536636353,
+      "learning_rate": 0.00033929999999999995,
+      "loss": 1.8404,
       "step": 1142
     },
     {
       "epoch": 1.8288,
-      "grad_norm": 1.3699275255203247,
-      "learning_rate": 0.00045239999999999994,
-      "loss": 1.9391,
+      "grad_norm": 1.6660796403884888,
+      "learning_rate": 0.00033959999999999996,
+      "loss": 1.3491,
       "step": 1143
     },
     {
       "epoch": 1.8304,
-      "grad_norm": 1.7976224422454834,
-      "learning_rate": 0.0004528,
-      "loss": 1.9632,
+      "grad_norm": 0.85002601146698,
+      "learning_rate": 0.00033989999999999997,
+      "loss": 1.2806,
       "step": 1144
     },
     {
       "epoch": 1.8319999999999999,
-      "grad_norm": 1.7157306671142578,
-      "learning_rate": 0.00045319999999999996,
-      "loss": 2.0015,
+      "grad_norm": 0.9421733617782593,
+      "learning_rate": 0.0003401999999999999,
+      "loss": 1.1968,
       "step": 1145
     },
     {
       "epoch": 1.8336000000000001,
-      "grad_norm": 1.1712660789489746,
-      "learning_rate": 0.00045359999999999997,
-      "loss": 1.7752,
+      "grad_norm": 0.7952862977981567,
+      "learning_rate": 0.0003405,
+      "loss": 1.4257,
       "step": 1146
     },
     {
       "epoch": 1.8352,
-      "grad_norm": 0.9054010510444641,
-      "learning_rate": 0.000454,
-      "loss": 2.0706,
+      "grad_norm": 1.2081334590911865,
+      "learning_rate": 0.00034079999999999994,
+      "loss": 1.3157,
       "step": 1147
     },
     {
       "epoch": 1.8368,
-      "grad_norm": 0.8177084922790527,
-      "learning_rate": 0.00045439999999999993,
-      "loss": 1.7316,
+      "grad_norm": 1.6554081439971924,
+      "learning_rate": 0.0003411,
+      "loss": 1.4017,
       "step": 1148
     },
     {
       "epoch": 1.8384,
-      "grad_norm": 1.8948848247528076,
-      "learning_rate": 0.00045479999999999994,
-      "loss": 1.9209,
+      "grad_norm": 1.1112024784088135,
+      "learning_rate": 0.00034139999999999995,
+      "loss": 1.3912,
       "step": 1149
     },
     {
       "epoch": 1.8399999999999999,
-      "grad_norm": 2.157670021057129,
-      "learning_rate": 0.0004552,
-      "loss": 1.9837,
+      "grad_norm": 1.1887140274047852,
+      "learning_rate": 0.00034169999999999996,
+      "loss": 1.4092,
       "step": 1150
     },
     {
       "epoch": 1.8416000000000001,
-      "grad_norm": 2.8265748023986816,
-      "learning_rate": 0.00045559999999999996,
-      "loss": 2.0989,
+      "grad_norm": 0.9006046056747437,
+      "learning_rate": 0.00034199999999999996,
+      "loss": 1.4782,
       "step": 1151
     },
     {
       "epoch": 1.8432,
-      "grad_norm": 2.864417552947998,
-      "learning_rate": 0.00045599999999999997,
-      "loss": 1.8949,
+      "grad_norm": 0.8354558944702148,
+      "learning_rate": 0.00034229999999999997,
+      "loss": 1.3304,
       "step": 1152
     },
     {
       "epoch": 1.8448,
-      "grad_norm": 2.235643148422241,
-      "learning_rate": 0.0004564,
-      "loss": 1.7172,
+      "grad_norm": 1.1101363897323608,
+      "learning_rate": 0.0003425999999999999,
+      "loss": 1.4787,
       "step": 1153
     },
     {
       "epoch": 1.8464,
-      "grad_norm": 1.4401240348815918,
-      "learning_rate": 0.00045679999999999994,
-      "loss": 1.9149,
+      "grad_norm": 0.8940262198448181,
+      "learning_rate": 0.0003429,
+      "loss": 1.4119,
       "step": 1154
     },
     {
       "epoch": 1.8479999999999999,
-      "grad_norm": 1.7679356336593628,
-      "learning_rate": 0.00045719999999999995,
-      "loss": 1.9761,
+      "grad_norm": 0.973086953163147,
+      "learning_rate": 0.00034319999999999994,
+      "loss": 1.207,
       "step": 1155
     },
     {
       "epoch": 1.8496000000000001,
-      "grad_norm": 0.6671059131622314,
-      "learning_rate": 0.0004576,
-      "loss": 1.7264,
+      "grad_norm": 0.8686078786849976,
+      "learning_rate": 0.0003435,
+      "loss": 1.41,
       "step": 1156
     },
     {
       "epoch": 1.8512,
-      "grad_norm": 2.028003215789795,
-      "learning_rate": 0.00045799999999999997,
-      "loss": 2.0366,
+      "grad_norm": 1.9869338274002075,
+      "learning_rate": 0.00034379999999999995,
+      "loss": 1.6477,
       "step": 1157
     },
     {
       "epoch": 1.8528,
-      "grad_norm": 1.4847235679626465,
-      "learning_rate": 0.0004584,
-      "loss": 1.8503,
+      "grad_norm": 1.961246371269226,
+      "learning_rate": 0.00034409999999999996,
+      "loss": 1.8708,
       "step": 1158
     },
     {
       "epoch": 1.8544,
-      "grad_norm": 1.177909016609192,
-      "learning_rate": 0.0004588,
-      "loss": 1.9129,
+      "grad_norm": 1.5795491933822632,
+      "learning_rate": 0.00034439999999999997,
+      "loss": 1.3292,
       "step": 1159
     },
     {
       "epoch": 1.8559999999999999,
-      "grad_norm": 0.9721006751060486,
-      "learning_rate": 0.00045919999999999994,
-      "loss": 1.8643,
+      "grad_norm": 1.0717681646347046,
+      "learning_rate": 0.0003447,
+      "loss": 1.127,
       "step": 1160
     },
     {
       "epoch": 1.8576000000000001,
-      "grad_norm": 2.1391115188598633,
-      "learning_rate": 0.00045959999999999995,
-      "loss": 2.0204,
+      "grad_norm": 1.2065831422805786,
+      "learning_rate": 0.00034499999999999993,
+      "loss": 1.3418,
       "step": 1161
     },
     {
       "epoch": 1.8592,
-      "grad_norm": 1.4983505010604858,
-      "learning_rate": 0.00046,
-      "loss": 1.7141,
+      "grad_norm": 1.1401423215866089,
+      "learning_rate": 0.0003453,
+      "loss": 1.596,
       "step": 1162
     },
     {
       "epoch": 1.8608,
-      "grad_norm": 1.2664161920547485,
-      "learning_rate": 0.00046039999999999997,
-      "loss": 1.7594,
+      "grad_norm": 1.329084873199463,
+      "learning_rate": 0.00034559999999999994,
+      "loss": 1.5699,
       "step": 1163
     },
     {
       "epoch": 1.8624,
-      "grad_norm": 0.9460088610649109,
-      "learning_rate": 0.0004608,
-      "loss": 1.6094,
+      "grad_norm": 1.0124589204788208,
+      "learning_rate": 0.00034589999999999995,
+      "loss": 1.4628,
       "step": 1164
     },
     {
       "epoch": 1.8639999999999999,
-      "grad_norm": 0.9295836091041565,
-      "learning_rate": 0.0004612,
-      "loss": 1.848,
+      "grad_norm": 1.0017642974853516,
+      "learning_rate": 0.00034619999999999996,
+      "loss": 1.3672,
       "step": 1165
     },
     {
       "epoch": 1.8656000000000001,
-      "grad_norm": 1.0059525966644287,
-      "learning_rate": 0.00046159999999999994,
-      "loss": 1.9431,
+      "grad_norm": 1.0151982307434082,
+      "learning_rate": 0.00034649999999999997,
+      "loss": 1.7449,
       "step": 1166
     },
     {
       "epoch": 1.8672,
-      "grad_norm": 0.6941124796867371,
-      "learning_rate": 0.00046199999999999995,
-      "loss": 1.5408,
+      "grad_norm": 0.9948499202728271,
+      "learning_rate": 0.0003467999999999999,
+      "loss": 1.9511,
       "step": 1167
     },
     {
       "epoch": 1.8688,
-      "grad_norm": 2.3069934844970703,
-      "learning_rate": 0.0004624,
-      "loss": 1.9049,
+      "grad_norm": 1.5604969263076782,
+      "learning_rate": 0.0003471,
+      "loss": 1.8143,
       "step": 1168
     },
     {
       "epoch": 1.8704,
-      "grad_norm": 1.7218568325042725,
-      "learning_rate": 0.0004628,
-      "loss": 1.634,
+      "grad_norm": NaN,
+      "learning_rate": 0.0003471,
+      "loss": 0.9545,
       "step": 1169
     },
     {
       "epoch": 1.8719999999999999,
-      "grad_norm": 0.6982250213623047,
-      "learning_rate": 0.0004632,
-      "loss": 2.0402,
+      "grad_norm": 1.3670181035995483,
+      "learning_rate": 0.00034739999999999993,
+      "loss": 2.0108,
       "step": 1170
     },
     {
       "epoch": 1.8736000000000002,
-      "grad_norm": 0.6958193778991699,
-      "learning_rate": 0.00046359999999999994,
-      "loss": 1.8532,
+      "grad_norm": 1.0110036134719849,
+      "learning_rate": 0.0003477,
+      "loss": 1.3934,
       "step": 1171
     },
     {
       "epoch": 1.8752,
-      "grad_norm": 3.401237964630127,
-      "learning_rate": 0.00046399999999999995,
-      "loss": 2.1863,
+      "grad_norm": 1.3402187824249268,
+      "learning_rate": 0.00034799999999999995,
+      "loss": 2.1356,
       "step": 1172
     },
     {
       "epoch": 1.8768,
-      "grad_norm": 1.5297185182571411,
-      "learning_rate": 0.00046439999999999996,
-      "loss": 2.0749,
+      "grad_norm": 1.2022168636322021,
+      "learning_rate": 0.00034829999999999996,
+      "loss": 2.0219,
       "step": 1173
     },
     {
       "epoch": 1.8784,
-      "grad_norm": 0.9487005472183228,
-      "learning_rate": 0.0004647999999999999,
-      "loss": 2.2366,
+      "grad_norm": 1.476257562637329,
+      "learning_rate": 0.00034859999999999996,
+      "loss": 1.9049,
       "step": 1174
     },
     {
       "epoch": 1.88,
-      "grad_norm": 1.712845802307129,
-      "learning_rate": 0.0004651999999999999,
-      "loss": 2.9485,
+      "grad_norm": NaN,
+      "learning_rate": 0.00034859999999999996,
+      "loss": 1.8173,
       "step": 1175
     },
     {
       "epoch": 1.8816000000000002,
-      "grad_norm": 8.302671432495117,
-      "learning_rate": 0.0004656,
-      "loss": 3.1518,
+      "grad_norm": 9.76489543914795,
+      "learning_rate": 0.00034889999999999997,
+      "loss": 2.8974,
       "step": 1176
     },
     {
       "epoch": 1.8832,
-      "grad_norm": 1.6076334714889526,
-      "learning_rate": 0.00046599999999999994,
-      "loss": 2.0009,
+      "grad_norm": 3.76676082611084,
+      "learning_rate": 0.0003491999999999999,
+      "loss": 1.9594,
       "step": 1177
     },
     {
       "epoch": 1.8848,
-      "grad_norm": 2.757471799850464,
-      "learning_rate": 0.00046639999999999995,
-      "loss": 1.8295,
+      "grad_norm": 3.224771022796631,
+      "learning_rate": 0.0003495,
+      "loss": 1.8728,
       "step": 1178
     },
     {
       "epoch": 1.8864,
-      "grad_norm": 2.844334602355957,
-      "learning_rate": 0.00046679999999999996,
-      "loss": 1.9377,
+      "grad_norm": 3.7804224491119385,
+      "learning_rate": 0.00034979999999999994,
+      "loss": 2.5339,
       "step": 1179
     },
     {
       "epoch": 1.888,
-      "grad_norm": 1.4853931665420532,
-      "learning_rate": 0.0004671999999999999,
-      "loss": 2.4697,
+      "grad_norm": 1.3428702354431152,
+      "learning_rate": 0.0003501,
+      "loss": 1.7693,
       "step": 1180
     },
     {
       "epoch": 1.8896,
-      "grad_norm": 5.221473693847656,
-      "learning_rate": 0.00046759999999999993,
-      "loss": 2.337,
+      "grad_norm": 1.6462411880493164,
+      "learning_rate": 0.00035039999999999995,
+      "loss": 1.9127,
       "step": 1181
     },
     {
       "epoch": 1.8912,
-      "grad_norm": 5.037121772766113,
-      "learning_rate": 0.000468,
-      "loss": 2.47,
+      "grad_norm": 3.412652015686035,
+      "learning_rate": 0.00035069999999999996,
+      "loss": 1.7551,
       "step": 1182
     },
     {
       "epoch": 1.8928,
-      "grad_norm": 4.587244510650635,
-      "learning_rate": 0.00046839999999999995,
-      "loss": 2.3116,
+      "grad_norm": 3.8241982460021973,
+      "learning_rate": 0.00035099999999999997,
+      "loss": 1.9243,
       "step": 1183
     },
     {
       "epoch": 1.8944,
-      "grad_norm": 3.243819236755371,
-      "learning_rate": 0.00046879999999999996,
-      "loss": 2.4598,
+      "grad_norm": 0.9238096475601196,
+      "learning_rate": 0.0003513,
+      "loss": 2.2319,
       "step": 1184
     },
     {
       "epoch": 1.896,
-      "grad_norm": 2.0243616104125977,
-      "learning_rate": 0.00046919999999999997,
-      "loss": 2.3673,
+      "grad_norm": 2.0135838985443115,
+      "learning_rate": 0.0003515999999999999,
+      "loss": 1.9178,
       "step": 1185
     },
     {
       "epoch": 1.8976,
-      "grad_norm": 0.9840572476387024,
-      "learning_rate": 0.0004695999999999999,
-      "loss": 2.2649,
+      "grad_norm": 2.6889097690582275,
+      "learning_rate": 0.0003519,
+      "loss": 1.8285,
       "step": 1186
     },
     {
       "epoch": 1.8992,
-      "grad_norm": 4.131119728088379,
-      "learning_rate": 0.00046999999999999993,
-      "loss": 2.5321,
+      "grad_norm": 1.5762139558792114,
+      "learning_rate": 0.00035219999999999994,
+      "loss": 1.9024,
       "step": 1187
     },
     {
       "epoch": 1.9008,
-      "grad_norm": 6.485975742340088,
-      "learning_rate": 0.0004704,
-      "loss": 2.8195,
+      "grad_norm": 0.8470468521118164,
+      "learning_rate": 0.0003525,
+      "loss": 1.7886,
       "step": 1188
     },
     {
       "epoch": 1.9024,
-      "grad_norm": 7.827768802642822,
-      "learning_rate": 0.00047079999999999995,
-      "loss": 2.8283,
+      "grad_norm": 1.188670039176941,
+      "learning_rate": 0.00035279999999999996,
+      "loss": 1.8531,
       "step": 1189
     },
     {
       "epoch": 1.904,
-      "grad_norm": 10.367870330810547,
-      "learning_rate": 0.00047119999999999996,
-      "loss": 3.0633,
+      "grad_norm": 2.7455601692199707,
+      "learning_rate": 0.00035309999999999996,
+      "loss": 1.9635,
       "step": 1190
     },
     {
       "epoch": 1.9056,
-      "grad_norm": 7.246518135070801,
-      "learning_rate": 0.00047159999999999997,
-      "loss": 2.6929,
+      "grad_norm": 3.5593600273132324,
+      "learning_rate": 0.00035339999999999997,
+      "loss": 1.9976,
       "step": 1191
     },
     {
       "epoch": 1.9072,
-      "grad_norm": 8.31074333190918,
-      "learning_rate": 0.0004719999999999999,
-      "loss": 2.9722,
+      "grad_norm": 4.159093379974365,
+      "learning_rate": 0.0003537,
+      "loss": 1.9367,
       "step": 1192
     },
     {
       "epoch": 1.9088,
-      "grad_norm": 13.296765327453613,
-      "learning_rate": 0.00047239999999999994,
-      "loss": 3.661,
+      "grad_norm": 4.765223026275635,
+      "learning_rate": 0.00035399999999999993,
+      "loss": 2.1654,
       "step": 1193
     },
     {
       "epoch": 1.9104,
-      "grad_norm": 5.0115580558776855,
-      "learning_rate": 0.0004728,
-      "loss": 2.7741,
+      "grad_norm": 6.879053115844727,
+      "learning_rate": 0.0003543,
+      "loss": 2.4834,
       "step": 1194
     },
     {
       "epoch": 1.912,
-      "grad_norm": 2.283982038497925,
-      "learning_rate": 0.00047319999999999996,
-      "loss": 2.4196,
+      "grad_norm": 5.095763683319092,
+      "learning_rate": 0.00035459999999999995,
+      "loss": 2.262,
       "step": 1195
     },
     {
       "epoch": 1.9136,
-      "grad_norm": 1.2423416376113892,
-      "learning_rate": 0.00047359999999999997,
-      "loss": 2.4451,
+      "grad_norm": 4.1628899574279785,
+      "learning_rate": 0.0003549,
+      "loss": 2.3063,
       "step": 1196
     },
     {
       "epoch": 1.9152,
-      "grad_norm": 1.4220050573349,
-      "learning_rate": 0.000474,
-      "loss": 2.5677,
+      "grad_norm": 2.7070200443267822,
+      "learning_rate": 0.00035519999999999996,
+      "loss": 2.1153,
       "step": 1197
     },
     {
       "epoch": 1.9167999999999998,
-      "grad_norm": 1.648946762084961,
-      "learning_rate": 0.00047439999999999993,
-      "loss": 2.5203,
+      "grad_norm": 2.1983084678649902,
+      "learning_rate": 0.00035549999999999997,
+      "loss": 1.8997,
       "step": 1198
     },
     {
       "epoch": 1.9184,
-      "grad_norm": 1.3786659240722656,
-      "learning_rate": 0.00047479999999999994,
-      "loss": 2.3582,
+      "grad_norm": 0.9442070126533508,
+      "learning_rate": 0.0003558,
+      "loss": 2.1675,
       "step": 1199
     },
     {
       "epoch": 1.92,
-      "grad_norm": 1.37655508518219,
-      "learning_rate": 0.0004752,
-      "loss": 2.5192,
-      "step": 1200
-    },
-    {
-      "epoch": 1.92,
-      "eval_cer": 0.6883644915693387,
-      "eval_loss": 2.5728604793548584,
-      "eval_runtime": 158.5376,
-      "eval_samples_per_second": 19.781,
-      "eval_steps_per_second": 1.236,
-      "eval_wer": 0.9177909024971506,
+      "grad_norm": 0.8357645869255066,
+      "learning_rate": 0.0003561,
+      "loss": 1.8516,
       "step": 1200
     },
     {
       "epoch": 1.9216,
-      "grad_norm": 1.187514066696167,
-      "learning_rate": 0.00047559999999999996,
-      "loss": 2.5463,
+      "grad_norm": 1.744762897491455,
+      "learning_rate": 0.00035639999999999994,
+      "loss": 2.0289,
       "step": 1201
     },
     {
       "epoch": 1.9232,
-      "grad_norm": 1.9947316646575928,
-      "learning_rate": 0.00047599999999999997,
-      "loss": 2.4633,
+      "grad_norm": 2.0870983600616455,
+      "learning_rate": 0.0003567,
+      "loss": 2.0565,
       "step": 1202
     },
     {
       "epoch": 1.9247999999999998,
-      "grad_norm": 3.0653204917907715,
-      "learning_rate": 0.0004764,
-      "loss": 2.6337,
+      "grad_norm": 1.1655802726745605,
+      "learning_rate": 0.00035699999999999995,
+      "loss": 2.0301,
       "step": 1203
     },
     {
       "epoch": 1.9264000000000001,
-      "grad_norm": 4.112390995025635,
-      "learning_rate": 0.00047679999999999993,
-      "loss": 2.4977,
+      "grad_norm": 1.394233226776123,
+      "learning_rate": 0.0003573,
+      "loss": 2.0413,
       "step": 1204
     },
     {
       "epoch": 1.928,
-      "grad_norm": 3.0475566387176514,
-      "learning_rate": 0.00047719999999999994,
-      "loss": 2.3692,
+      "grad_norm": 0.8284745216369629,
+      "learning_rate": 0.00035759999999999996,
+      "loss": 2.0316,
       "step": 1205
     },
     {
       "epoch": 1.9296,
-      "grad_norm": 3.443176507949829,
-      "learning_rate": 0.0004776,
-      "loss": 2.5523,
+      "grad_norm": 1.0152744054794312,
+      "learning_rate": 0.00035789999999999997,
+      "loss": 2.0509,
       "step": 1206
     },
     {
       "epoch": 1.9312,
-      "grad_norm": 2.9816811084747314,
-      "learning_rate": 0.00047799999999999996,
-      "loss": 2.5454,
+      "grad_norm": 1.3899942636489868,
+      "learning_rate": 0.0003582,
+      "loss": 1.8778,
       "step": 1207
     },
     {
       "epoch": 1.9327999999999999,
-      "grad_norm": 2.6588003635406494,
-      "learning_rate": 0.0004784,
-      "loss": 2.3668,
+      "grad_norm": 2.3678336143493652,
+      "learning_rate": 0.0003585,
+      "loss": 2.128,
       "step": 1208
     },
     {
       "epoch": 1.9344000000000001,
-      "grad_norm": 0.9880726933479309,
-      "learning_rate": 0.0004788,
-      "loss": 2.5404,
+      "grad_norm": 2.9605188369750977,
+      "learning_rate": 0.00035879999999999994,
+      "loss": 1.902,
       "step": 1209
     },
     {
       "epoch": 1.936,
-      "grad_norm": 1.2105108499526978,
-      "learning_rate": 0.00047919999999999994,
-      "loss": 2.5146,
+      "grad_norm": 1.4265320301055908,
+      "learning_rate": 0.0003591,
+      "loss": 2.0587,
       "step": 1210
     },
     {
       "epoch": 1.9376,
-      "grad_norm": 0.7539094090461731,
-      "learning_rate": 0.00047959999999999995,
-      "loss": 2.5409,
+      "grad_norm": 2.7985453605651855,
+      "learning_rate": 0.00035939999999999995,
+      "loss": 2.1007,
       "step": 1211
     },
     {
       "epoch": 1.9392,
-      "grad_norm": 1.3209741115570068,
-      "learning_rate": 0.00047999999999999996,
-      "loss": 2.2839,
+      "grad_norm": 1.6142802238464355,
+      "learning_rate": 0.00035969999999999996,
+      "loss": 1.7844,
       "step": 1212
     },
     {
       "epoch": 1.9407999999999999,
-      "grad_norm": 0.5982227325439453,
-      "learning_rate": 0.0004803999999999999,
-      "loss": 2.3925,
+      "grad_norm": 0.7082346081733704,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 1.9045,
       "step": 1213
     },
     {
       "epoch": 1.9424000000000001,
-      "grad_norm": 1.774712085723877,
-      "learning_rate": 0.0004808,
-      "loss": 2.0152,
+      "grad_norm": 0.6736663579940796,
+      "learning_rate": 0.0003603,
+      "loss": 1.9604,
       "step": 1214
     },
     {
       "epoch": 1.944,
-      "grad_norm": 1.231215238571167,
-      "learning_rate": 0.0004812,
-      "loss": 2.271,
+      "grad_norm": 0.8059405088424683,
+      "learning_rate": 0.00036059999999999993,
+      "loss": 1.8783,
       "step": 1215
     },
     {
       "epoch": 1.9456,
-      "grad_norm": 1.5429651737213135,
-      "learning_rate": 0.00048159999999999994,
-      "loss": 2.4412,
+      "grad_norm": 0.795032262802124,
+      "learning_rate": 0.0003609,
+      "loss": 1.6925,
       "step": 1216
     },
     {
       "epoch": 1.9472,
-      "grad_norm": 1.844327449798584,
-      "learning_rate": 0.00048199999999999995,
-      "loss": 2.5497,
+      "grad_norm": 0.7345133423805237,
+      "learning_rate": 0.00036119999999999994,
+      "loss": 1.6741,
       "step": 1217
     },
     {
       "epoch": 1.9487999999999999,
-      "grad_norm": 3.4878883361816406,
-      "learning_rate": 0.00048239999999999996,
-      "loss": 2.2781,
+      "grad_norm": 1.4538378715515137,
+      "learning_rate": 0.0003615,
+      "loss": 1.6525,
       "step": 1218
     },
     {
       "epoch": 1.9504000000000001,
-      "grad_norm": 0.906635046005249,
-      "learning_rate": 0.0004827999999999999,
-      "loss": 2.3062,
+      "grad_norm": 2.7808940410614014,
+      "learning_rate": 0.00036179999999999996,
+      "loss": 1.73,
       "step": 1219
     },
     {
       "epoch": 1.952,
-      "grad_norm": 1.1018857955932617,
-      "learning_rate": 0.0004832,
-      "loss": 2.2925,
+      "grad_norm": 1.7963067293167114,
+      "learning_rate": 0.00036209999999999997,
+      "loss": 1.8387,
       "step": 1220
     },
     {
       "epoch": 1.9536,
-      "grad_norm": 0.7406502366065979,
-      "learning_rate": 0.0004836,
-      "loss": 2.0683,
+      "grad_norm": 1.8159714937210083,
+      "learning_rate": 0.00036239999999999997,
+      "loss": 2.1202,
       "step": 1221
     },
     {
       "epoch": 1.9552,
-      "grad_norm": 1.1070985794067383,
-      "learning_rate": 0.00048399999999999995,
-      "loss": 2.2204,
+      "grad_norm": 1.3005350828170776,
+      "learning_rate": 0.0003627,
+      "loss": 2.1657,
       "step": 1222
     },
     {
       "epoch": 1.9567999999999999,
-      "grad_norm": 1.2866196632385254,
-      "learning_rate": 0.00048439999999999996,
-      "loss": 2.5341,
+      "grad_norm": 0.9724456071853638,
+      "learning_rate": 0.00036299999999999993,
+      "loss": 2.1556,
       "step": 1223
     },
     {
       "epoch": 1.9584000000000001,
-      "grad_norm": 2.646087408065796,
-      "learning_rate": 0.00048479999999999997,
-      "loss": 2.4654,
+      "grad_norm": 0.9316180944442749,
+      "learning_rate": 0.0003633,
+      "loss": 1.5908,
       "step": 1224
     },
     {
       "epoch": 1.96,
-      "grad_norm": 2.4793953895568848,
-      "learning_rate": 0.0004851999999999999,
-      "loss": 2.4107,
+      "grad_norm": 1.7360824346542358,
+      "learning_rate": 0.00036359999999999995,
+      "loss": 2.3979,
       "step": 1225
     },
     {
       "epoch": 1.9616,
-      "grad_norm": 9.734627723693848,
-      "learning_rate": 0.0004856,
-      "loss": 2.7599,
+      "grad_norm": 17.20070457458496,
+      "learning_rate": 0.0003639,
+      "loss": 3.8859,
       "step": 1226
     },
     {
       "epoch": 1.9632,
-      "grad_norm": 14.538433074951172,
-      "learning_rate": 0.000486,
-      "loss": 3.1676,
+      "grad_norm": 7.804291248321533,
+      "learning_rate": 0.00036419999999999996,
+      "loss": 2.8236,
       "step": 1227
     },
     {
       "epoch": 1.9647999999999999,
-      "grad_norm": 1.0963908433914185,
-      "learning_rate": 0.00048639999999999995,
-      "loss": 2.0805,
+      "grad_norm": 4.541631698608398,
+      "learning_rate": 0.00036449999999999997,
+      "loss": 2.2704,
       "step": 1228
     },
     {
       "epoch": 1.9664000000000001,
-      "grad_norm": 2.8844449520111084,
-      "learning_rate": 0.00048679999999999996,
-      "loss": 2.1634,
+      "grad_norm": 5.615588665008545,
+      "learning_rate": 0.0003648,
+      "loss": 2.4331,
       "step": 1229
     },
     {
       "epoch": 1.968,
-      "grad_norm": 1.3918949365615845,
-      "learning_rate": 0.00048719999999999997,
-      "loss": 2.1095,
+      "grad_norm": 2.0212814807891846,
+      "learning_rate": 0.0003651,
+      "loss": 2.0981,
       "step": 1230
     },
     {
       "epoch": 1.9696,
-      "grad_norm": 1.6348989009857178,
-      "learning_rate": 0.0004875999999999999,
-      "loss": 1.9739,
+      "grad_norm": 2.8328936100006104,
+      "learning_rate": 0.00036539999999999994,
+      "loss": 1.9869,
       "step": 1231
     },
     {
       "epoch": 1.9712,
-      "grad_norm": 3.712069034576416,
-      "learning_rate": 0.000488,
-      "loss": 1.8339,
+      "grad_norm": 3.4782793521881104,
+      "learning_rate": 0.0003657,
+      "loss": 2.05,
       "step": 1232
     },
     {
       "epoch": 1.9727999999999999,
-      "grad_norm": 3.0852272510528564,
-      "learning_rate": 0.0004883999999999999,
-      "loss": 1.7957,
+      "grad_norm": 2.7648117542266846,
+      "learning_rate": 0.00036599999999999995,
+      "loss": 2.0339,
       "step": 1233
     },
     {
       "epoch": 1.9744000000000002,
-      "grad_norm": 2.9243290424346924,
-      "learning_rate": 0.0004888,
-      "loss": 1.9496,
+      "grad_norm": 3.1286776065826416,
+      "learning_rate": 0.0003663,
+      "loss": 2.0052,
       "step": 1234
     },
     {
       "epoch": 1.976,
-      "grad_norm": 2.6891889572143555,
-      "learning_rate": 0.0004892,
-      "loss": 1.704,
+      "grad_norm": 1.4266185760498047,
+      "learning_rate": 0.00036659999999999997,
+      "loss": 2.1165,
       "step": 1235
     },
     {
       "epoch": 1.9776,
-      "grad_norm": 1.1795345544815063,
-      "learning_rate": 0.0004896,
-      "loss": 2.0643,
+      "grad_norm": 1.3429757356643677,
+      "learning_rate": 0.0003669,
+      "loss": 2.033,
       "step": 1236
     },
     {
       "epoch": 1.9792,
-      "grad_norm": 2.66219425201416,
-      "learning_rate": 0.00049,
-      "loss": 2.0074,
+      "grad_norm": 0.6780145168304443,
+      "learning_rate": 0.0003672,
+      "loss": 2.0324,
       "step": 1237
     },
     {
       "epoch": 1.9808,
-      "grad_norm": 3.1762375831604004,
-      "learning_rate": 0.0004904,
-      "loss": 1.9707,
+      "grad_norm": 2.1947433948516846,
+      "learning_rate": 0.0003675,
+      "loss": 2.0592,
       "step": 1238
     },
     {
       "epoch": 1.9824000000000002,
-      "grad_norm": 6.450320243835449,
-      "learning_rate": 0.0004907999999999999,
-      "loss": 2.1798,
+      "grad_norm": 4.260788440704346,
+      "learning_rate": 0.00036779999999999994,
+      "loss": 2.0549,
       "step": 1239
     },
     {
       "epoch": 1.984,
-      "grad_norm": 2.975370168685913,
-      "learning_rate": 0.0004911999999999999,
-      "loss": 2.0042,
+      "grad_norm": 5.300332069396973,
+      "learning_rate": 0.0003681,
+      "loss": 2.2587,
       "step": 1240
     },
     {
       "epoch": 1.9856,
-      "grad_norm": 4.335202217102051,
-      "learning_rate": 0.0004916,
-      "loss": 2.0929,
+      "grad_norm": 6.130112171173096,
+      "learning_rate": 0.00036839999999999996,
+      "loss": 2.3789,
       "step": 1241
     },
     {
       "epoch": 1.9872,
-      "grad_norm": 3.7740931510925293,
-      "learning_rate": 0.0004919999999999999,
-      "loss": 2.2352,
+      "grad_norm": 7.301811218261719,
+      "learning_rate": 0.0003687,
+      "loss": 2.543,
       "step": 1242
     },
     {
       "epoch": 1.9888,
-      "grad_norm": 1.6590664386749268,
-      "learning_rate": 0.0004923999999999999,
-      "loss": 2.1172,
+      "grad_norm": 5.791845321655273,
+      "learning_rate": 0.00036899999999999997,
+      "loss": 2.2248,
       "step": 1243
     },
     {
       "epoch": 1.9904,
-      "grad_norm": 1.345360279083252,
-      "learning_rate": 0.0004927999999999999,
-      "loss": 1.9999,
+      "grad_norm": 5.044650077819824,
+      "learning_rate": 0.0003693,
+      "loss": 2.3361,
       "step": 1244
     },
     {
       "epoch": 1.992,
-      "grad_norm": 0.9986281991004944,
-      "learning_rate": 0.0004932,
-      "loss": 1.9315,
+      "grad_norm": 4.885112285614014,
+      "learning_rate": 0.0003696,
+      "loss": 2.3957,
       "step": 1245
     },
     {
       "epoch": 1.9936,
-      "grad_norm": 1.795646071434021,
-      "learning_rate": 0.0004936,
-      "loss": 2.275,
+      "grad_norm": 2.838794469833374,
+      "learning_rate": 0.0003699,
+      "loss": 2.2255,
       "step": 1246
     },
     {
       "epoch": 1.9952,
-      "grad_norm": 1.1249878406524658,
-      "learning_rate": 0.000494,
-      "loss": 2.2124,
+      "grad_norm": 2.508756637573242,
+      "learning_rate": 0.00037019999999999995,
+      "loss": 2.1271,
       "step": 1247
     },
     {
       "epoch": 1.9968,
-      "grad_norm": 0.8439085483551025,
-      "learning_rate": 0.0004944,
-      "loss": 2.0854,
+      "grad_norm": 1.2168052196502686,
+      "learning_rate": 0.0003705,
+      "loss": 2.1874,
       "step": 1248
     },
     {
       "epoch": 1.9984,
-      "grad_norm": 1.0127602815628052,
-      "learning_rate": 0.0004948,
-      "loss": 2.3189,
+      "grad_norm": 2.0647408962249756,
+      "learning_rate": 0.00037079999999999996,
+      "loss": 2.3272,
       "step": 1249
     },
     {
       "epoch": 2.0,
-      "grad_norm": 1.2832350730895996,
-      "learning_rate": 0.0004952,
-      "loss": 2.3423,
+      "grad_norm": 2.734804630279541,
+      "learning_rate": 0.0003711,
+      "loss": 2.3636,
       "step": 1250
     },
     {
       "epoch": 2.0016,
-      "grad_norm": 16.80681800842285,
-      "learning_rate": 0.0004955999999999999,
-      "loss": 3.3692,
+      "grad_norm": 1.4602904319763184,
+      "learning_rate": 0.0003714,
+      "loss": 2.7633,
       "step": 1251
     },
     {
       "epoch": 2.0032,
-      "grad_norm": 11.572160720825195,
-      "learning_rate": 0.0004959999999999999,
-      "loss": 2.8517,
+      "grad_norm": 0.7962429523468018,
+      "learning_rate": 0.0003717,
+      "loss": 2.4137,
       "step": 1252
     },
     {
       "epoch": 2.0048,
-      "grad_norm": 8.817554473876953,
-      "learning_rate": 0.0004963999999999999,
-      "loss": 2.5597,
+      "grad_norm": 2.490060806274414,
+      "learning_rate": 0.000372,
+      "loss": 2.5798,
       "step": 1253
     },
     {
       "epoch": 2.0064,
-      "grad_norm": 5.953640937805176,
-      "learning_rate": 0.0004967999999999999,
-      "loss": 2.3292,
+      "grad_norm": 3.1522207260131836,
+      "learning_rate": 0.0003723,
+      "loss": 2.3581,
       "step": 1254
     },
     {
       "epoch": 2.008,
-      "grad_norm": 6.86954927444458,
-      "learning_rate": 0.0004971999999999999,
-      "loss": 2.6954,
+      "grad_norm": 1.5471434593200684,
+      "learning_rate": 0.00037259999999999995,
+      "loss": 2.7235,
       "step": 1255
     },
     {
       "epoch": 2.0096,
-      "grad_norm": 1.5113718509674072,
-      "learning_rate": 0.0004976,
-      "loss": 2.0988,
+      "grad_norm": 4.790720462799072,
+      "learning_rate": 0.0003729,
+      "loss": 2.424,
       "step": 1256
     },
     {
       "epoch": 2.0112,
-      "grad_norm": 1.745615005493164,
-      "learning_rate": 0.000498,
-      "loss": 1.985,
+      "grad_norm": 3.654318332672119,
+      "learning_rate": 0.00037319999999999996,
+      "loss": 2.4581,
       "step": 1257
     },
     {
       "epoch": 2.0128,
-      "grad_norm": 2.888578414916992,
-      "learning_rate": 0.0004984,
-      "loss": 1.9931,
+      "grad_norm": 4.856836318969727,
+      "learning_rate": 0.0003735,
+      "loss": 2.4449,
       "step": 1258
     },
     {
       "epoch": 2.0144,
-      "grad_norm": 2.6909291744232178,
-      "learning_rate": 0.0004988,
-      "loss": 2.3251,
+      "grad_norm": 3.019021987915039,
+      "learning_rate": 0.0003738,
+      "loss": 2.56,
       "step": 1259
     },
     {
       "epoch": 2.016,
-      "grad_norm": 4.514300346374512,
-      "learning_rate": 0.0004991999999999999,
-      "loss": 2.1243,
+      "grad_norm": 2.4978408813476562,
+      "learning_rate": 0.0003741,
+      "loss": 2.2926,
       "step": 1260
     },
     {
       "epoch": 2.0176,
-      "grad_norm": 2.5229086875915527,
-      "learning_rate": 0.0004996,
-      "loss": 2.2339,
+      "grad_norm": 1.7391501665115356,
+      "learning_rate": 0.0003744,
+      "loss": 2.5892,
       "step": 1261
     },
     {
       "epoch": 2.0192,
-      "grad_norm": 5.69707727432251,
-      "learning_rate": 0.0005,
-      "loss": 2.167,
+      "grad_norm": 1.032046914100647,
+      "learning_rate": 0.0003747,
+      "loss": 2.2562,
       "step": 1262
     },
     {
       "epoch": 2.0208,
-      "grad_norm": 4.998229503631592,
-      "learning_rate": 0.0005003999999999999,
-      "loss": 2.3208,
+      "grad_norm": 1.8993191719055176,
+      "learning_rate": 0.00037499999999999995,
+      "loss": 2.2962,
       "step": 1263
     },
     {
       "epoch": 2.0224,
-      "grad_norm": 5.518265247344971,
-      "learning_rate": 0.0005007999999999999,
-      "loss": 2.1296,
+      "grad_norm": 5.481810569763184,
+      "learning_rate": 0.00037529999999999996,
+      "loss": 2.4923,
       "step": 1264
     },
     {
       "epoch": 2.024,
-      "grad_norm": 5.083476543426514,
-      "learning_rate": 0.0005011999999999999,
-      "loss": 2.3207,
+      "grad_norm": 6.432821273803711,
+      "learning_rate": 0.00037559999999999997,
+      "loss": 2.4889,
       "step": 1265
     },
     {
       "epoch": 2.0256,
-      "grad_norm": 4.121245861053467,
-      "learning_rate": 0.0005015999999999999,
-      "loss": 2.0558,
+      "grad_norm": 7.250077724456787,
+      "learning_rate": 0.0003758999999999999,
+      "loss": 2.5924,
       "step": 1266
     },
     {
       "epoch": 2.0272,
-      "grad_norm": 2.552358865737915,
-      "learning_rate": 0.000502,
-      "loss": 2.1283,
+      "grad_norm": 7.486959457397461,
+      "learning_rate": 0.0003762,
+      "loss": 2.5582,
       "step": 1267
     },
     {
       "epoch": 2.0288,
-      "grad_norm": 1.2170063257217407,
-      "learning_rate": 0.0005024,
-      "loss": 2.1844,
+      "grad_norm": 8.361715316772461,
+      "learning_rate": 0.00037649999999999994,
+      "loss": 2.6335,
       "step": 1268
     },
     {
       "epoch": 2.0304,
-      "grad_norm": 1.5925713777542114,
-      "learning_rate": 0.0005028,
-      "loss": 2.1698,
+      "grad_norm": 6.474916934967041,
+      "learning_rate": 0.00037679999999999994,
+      "loss": 2.4855,
       "step": 1269
     },
     {
       "epoch": 2.032,
-      "grad_norm": 4.9720377922058105,
-      "learning_rate": 0.0005032,
-      "loss": 2.2092,
+      "grad_norm": 7.403282165527344,
+      "learning_rate": 0.00037709999999999995,
+      "loss": 2.4993,
       "step": 1270
     },
     {
       "epoch": 2.0336,
-      "grad_norm": 5.816512107849121,
-      "learning_rate": 0.0005036,
-      "loss": 2.5182,
+      "grad_norm": 5.777534008026123,
+      "learning_rate": 0.00037739999999999996,
+      "loss": 2.4794,
       "step": 1271
     },
     {
       "epoch": 2.0352,
-      "grad_norm": 4.63362455368042,
-      "learning_rate": 0.0005039999999999999,
-      "loss": 2.127,
+      "grad_norm": 4.98473596572876,
+      "learning_rate": 0.0003776999999999999,
+      "loss": 2.3016,
       "step": 1272
     },
     {
       "epoch": 2.0368,
-      "grad_norm": 4.630200386047363,
-      "learning_rate": 0.0005044,
-      "loss": 2.296,
+      "grad_norm": 5.620299339294434,
+      "learning_rate": 0.00037799999999999997,
+      "loss": 2.4113,
       "step": 1273
     },
     {
       "epoch": 2.0384,
-      "grad_norm": 5.835329532623291,
-      "learning_rate": 0.0005048,
-      "loss": 2.2678,
+      "grad_norm": 3.0978798866271973,
+      "learning_rate": 0.0003782999999999999,
+      "loss": 2.3554,
       "step": 1274
     },
     {
       "epoch": 2.04,
-      "grad_norm": 2.8462319374084473,
-      "learning_rate": 0.0005051999999999999,
-      "loss": 2.1091,
+      "grad_norm": 1.7908412218093872,
+      "learning_rate": 0.0003786,
+      "loss": 2.2717,
       "step": 1275
     },
     {
       "epoch": 2.0416,
-      "grad_norm": 2.3021416664123535,
-      "learning_rate": 0.0005055999999999999,
-      "loss": 2.212,
+      "grad_norm": 1.7431128025054932,
+      "learning_rate": 0.00037889999999999994,
+      "loss": 2.114,
       "step": 1276
     },
     {
       "epoch": 2.0432,
-      "grad_norm": 1.7580206394195557,
-      "learning_rate": 0.0005059999999999999,
-      "loss": 2.2206,
+      "grad_norm": 1.9171562194824219,
+      "learning_rate": 0.00037919999999999995,
+      "loss": 2.3205,
       "step": 1277
     },
     {
       "epoch": 2.0448,
-      "grad_norm": 0.9904365539550781,
-      "learning_rate": 0.0005064,
-      "loss": 2.2598,
+      "grad_norm": 1.3162221908569336,
+      "learning_rate": 0.00037949999999999995,
+      "loss": 2.1851,
       "step": 1278
     },
     {
       "epoch": 2.0464,
-      "grad_norm": 1.5216723680496216,
-      "learning_rate": 0.0005068,
-      "loss": 2.0426,
+      "grad_norm": 1.1935993432998657,
+      "learning_rate": 0.00037979999999999996,
+      "loss": 2.1238,
       "step": 1279
     },
     {
       "epoch": 2.048,
-      "grad_norm": 0.9763423204421997,
-      "learning_rate": 0.0005072,
-      "loss": 2.2838,
+      "grad_norm": 1.006394624710083,
+      "learning_rate": 0.0003800999999999999,
+      "loss": 2.1991,
       "step": 1280
     },
     {
       "epoch": 2.0496,
-      "grad_norm": 0.9675254821777344,
-      "learning_rate": 0.0005076,
-      "loss": 2.2737,
+      "grad_norm": 0.6535553336143494,
+      "learning_rate": 0.0003804,
+      "loss": 2.2614,
       "step": 1281
     },
     {
       "epoch": 2.0512,
-      "grad_norm": 0.874781608581543,
-      "learning_rate": 0.000508,
-      "loss": 2.0927,
+      "grad_norm": 2.4608588218688965,
+      "learning_rate": 0.00038069999999999993,
+      "loss": 1.9734,
       "step": 1282
     },
     {
       "epoch": 2.0528,
-      "grad_norm": 1.92698335647583,
-      "learning_rate": 0.0005084,
-      "loss": 2.3211,
+      "grad_norm": 2.3777921199798584,
+      "learning_rate": 0.000381,
+      "loss": 1.9419,
       "step": 1283
     },
     {
       "epoch": 2.0544,
-      "grad_norm": 1.462214708328247,
-      "learning_rate": 0.0005087999999999999,
-      "loss": 2.0673,
+      "grad_norm": 3.6456470489501953,
+      "learning_rate": 0.00038129999999999994,
+      "loss": 2.1727,
       "step": 1284
     },
     {
       "epoch": 2.056,
-      "grad_norm": 1.4558961391448975,
-      "learning_rate": 0.0005092,
-      "loss": 2.4106,
+      "grad_norm": 3.8108816146850586,
+      "learning_rate": 0.00038159999999999995,
+      "loss": 2.244,
       "step": 1285
     },
     {
       "epoch": 2.0576,
-      "grad_norm": 1.2148739099502563,
-      "learning_rate": 0.0005096,
-      "loss": 2.3889,
+      "grad_norm": 3.471174478530884,
+      "learning_rate": 0.00038189999999999996,
+      "loss": 2.2731,
       "step": 1286
     },
     {
       "epoch": 2.0592,
-      "grad_norm": 1.752792239189148,
-      "learning_rate": 0.0005099999999999999,
-      "loss": 2.3579,
+      "grad_norm": 1.8833621740341187,
+      "learning_rate": 0.00038219999999999997,
+      "loss": 2.3379,
       "step": 1287
     },
     {
       "epoch": 2.0608,
-      "grad_norm": 0.8984630107879639,
-      "learning_rate": 0.0005103999999999999,
-      "loss": 2.3451,
+      "grad_norm": 2.832181453704834,
+      "learning_rate": 0.0003824999999999999,
+      "loss": 2.064,
       "step": 1288
     },
     {
       "epoch": 2.0624,
-      "grad_norm": 0.6344359517097473,
-      "learning_rate": 0.0005108,
-      "loss": 2.2167,
+      "grad_norm": 1.3994871377944946,
+      "learning_rate": 0.0003828,
+      "loss": 2.0979,
       "step": 1289
     },
     {
       "epoch": 2.064,
-      "grad_norm": 0.4872526526451111,
-      "learning_rate": 0.0005112,
-      "loss": 2.435,
+      "grad_norm": 0.7377781271934509,
+      "learning_rate": 0.00038309999999999993,
+      "loss": 2.127,
       "step": 1290
     },
     {
       "epoch": 2.0656,
-      "grad_norm": 1.0829209089279175,
-      "learning_rate": 0.0005116,
-      "loss": 2.2084,
+      "grad_norm": 1.0238137245178223,
+      "learning_rate": 0.0003834,
+      "loss": 2.0696,
       "step": 1291
     },
     {
       "epoch": 2.0672,
-      "grad_norm": 1.2053056955337524,
-      "learning_rate": 0.000512,
-      "loss": 2.273,
+      "grad_norm": 1.6712844371795654,
+      "learning_rate": 0.00038369999999999995,
+      "loss": 2.2395,
       "step": 1292
     },
     {
       "epoch": 2.0688,
-      "grad_norm": 2.491065263748169,
-      "learning_rate": 0.0005124,
-      "loss": 2.5486,
+      "grad_norm": 0.8299332857131958,
+      "learning_rate": 0.00038399999999999996,
+      "loss": 2.2809,
       "step": 1293
     },
     {
       "epoch": 2.0704,
-      "grad_norm": 0.7429988384246826,
-      "learning_rate": 0.0005128,
-      "loss": 2.4695,
+      "grad_norm": 1.4740967750549316,
+      "learning_rate": 0.00038429999999999996,
+      "loss": 2.2029,
       "step": 1294
     },
     {
       "epoch": 2.072,
-      "grad_norm": 1.1446125507354736,
-      "learning_rate": 0.0005131999999999999,
-      "loss": 2.3198,
+      "grad_norm": 1.6210805177688599,
+      "learning_rate": 0.00038459999999999997,
+      "loss": 2.1513,
       "step": 1295
     },
     {
       "epoch": 2.0736,
-      "grad_norm": 0.8218686580657959,
-      "learning_rate": 0.0005135999999999999,
-      "loss": 2.3743,
+      "grad_norm": 0.7904883027076721,
+      "learning_rate": 0.0003848999999999999,
+      "loss": 2.0126,
       "step": 1296
     },
     {
       "epoch": 2.0752,
-      "grad_norm": 1.084084153175354,
-      "learning_rate": 0.0005139999999999999,
-      "loss": 2.4759,
+      "grad_norm": 1.0255601406097412,
+      "learning_rate": 0.0003852,
+      "loss": 2.0117,
       "step": 1297
     },
     {
       "epoch": 2.0768,
-      "grad_norm": 1.4052973985671997,
-      "learning_rate": 0.0005143999999999999,
-      "loss": 2.3302,
+      "grad_norm": 2.568159818649292,
+      "learning_rate": 0.00038549999999999994,
+      "loss": 1.9919,
       "step": 1298
     },
     {
       "epoch": 2.0784,
-      "grad_norm": 5.035649299621582,
-      "learning_rate": 0.0005147999999999999,
-      "loss": 2.3812,
+      "grad_norm": 0.7457689046859741,
+      "learning_rate": 0.0003858,
+      "loss": 1.8096,
       "step": 1299
     },
     {
       "epoch": 2.08,
-      "grad_norm": 1.4146735668182373,
-      "learning_rate": 0.0005152,
-      "loss": 2.6077,
-      "step": 1300
-    },
-    {
-      "epoch": 2.08,
-      "eval_cer": 0.4815960457607526,
-      "eval_loss": 2.4077656269073486,
-      "eval_runtime": 158.706,
-      "eval_samples_per_second": 19.76,
-      "eval_steps_per_second": 1.235,
-      "eval_wer": 0.8066314371567713,
+      "grad_norm": 2.0637600421905518,
+      "learning_rate": 0.00038609999999999995,
+      "loss": 2.628,
       "step": 1300
     },
     {
       "epoch": 2.0816,
-      "grad_norm": 15.680081367492676,
-      "learning_rate": 0.0005156,
-      "loss": 3.1347,
+      "grad_norm": 7.795231819152832,
+      "learning_rate": 0.00038639999999999996,
+      "loss": 2.7336,
       "step": 1301
     },
     {
       "epoch": 2.0832,
-      "grad_norm": 7.167609214782715,
-      "learning_rate": 0.000516,
-      "loss": 2.594,
+      "grad_norm": 2.4228038787841797,
+      "learning_rate": 0.00038669999999999997,
+      "loss": 2.4832,
       "step": 1302
     },
     {
       "epoch": 2.0848,
-      "grad_norm": 2.3701157569885254,
-      "learning_rate": 0.0005164,
-      "loss": 2.4183,
+      "grad_norm": 7.511104583740234,
+      "learning_rate": 0.000387,
+      "loss": 3.1255,
       "step": 1303
     },
     {
       "epoch": 2.0864,
-      "grad_norm": 7.440915107727051,
-      "learning_rate": 0.0005167999999999999,
-      "loss": 2.8736,
+      "grad_norm": 4.0060882568359375,
+      "learning_rate": 0.00038729999999999993,
+      "loss": 2.4608,
       "step": 1304
     },
     {
       "epoch": 2.088,
-      "grad_norm": 2.976489543914795,
-      "learning_rate": 0.0005172,
-      "loss": 2.5311,
+      "grad_norm": 6.4439496994018555,
+      "learning_rate": 0.0003876,
+      "loss": 2.6808,
       "step": 1305
     },
     {
       "epoch": 2.0896,
-      "grad_norm": 4.291921615600586,
-      "learning_rate": 0.0005176,
-      "loss": 2.2646,
+      "grad_norm": 7.643751621246338,
+      "learning_rate": 0.00038789999999999994,
+      "loss": 2.8287,
       "step": 1306
     },
     {
       "epoch": 2.0912,
-      "grad_norm": 8.423929214477539,
-      "learning_rate": 0.0005179999999999999,
-      "loss": 2.3763,
+      "grad_norm": 7.657688140869141,
+      "learning_rate": 0.0003882,
+      "loss": 2.7325,
       "step": 1307
     },
     {
       "epoch": 2.0928,
-      "grad_norm": 8.711053848266602,
-      "learning_rate": 0.0005183999999999999,
-      "loss": 2.5501,
+      "grad_norm": 8.446870803833008,
+      "learning_rate": 0.00038849999999999996,
+      "loss": 3.022,
       "step": 1308
     },
     {
       "epoch": 2.0944,
-      "grad_norm": 8.829360008239746,
-      "learning_rate": 0.0005187999999999999,
-      "loss": 2.3476,
+      "grad_norm": 8.167088508605957,
+      "learning_rate": 0.00038879999999999996,
+      "loss": 3.0802,
       "step": 1309
     },
     {
       "epoch": 2.096,
-      "grad_norm": 8.021748542785645,
-      "learning_rate": 0.0005191999999999999,
-      "loss": 2.4302,
+      "grad_norm": 7.480300426483154,
+      "learning_rate": 0.00038909999999999997,
+      "loss": 2.8137,
       "step": 1310
     },
     {
       "epoch": 2.0976,
-      "grad_norm": 3.774869918823242,
-      "learning_rate": 0.0005195999999999999,
-      "loss": 2.4614,
+      "grad_norm": 6.842907905578613,
+      "learning_rate": 0.0003894,
+      "loss": 2.7077,
       "step": 1311
     },
     {
       "epoch": 2.0992,
-      "grad_norm": 6.809609413146973,
-      "learning_rate": 0.00052,
-      "loss": 2.2534,
+      "grad_norm": 5.104220867156982,
+      "learning_rate": 0.00038969999999999993,
+      "loss": 2.4495,
       "step": 1312
     },
     {
       "epoch": 2.1008,
-      "grad_norm": 4.694296836853027,
-      "learning_rate": 0.0005204,
-      "loss": 2.1121,
+      "grad_norm": 3.758356809616089,
+      "learning_rate": 0.00039,
+      "loss": 2.239,
       "step": 1313
     },
     {
       "epoch": 2.1024,
-      "grad_norm": 4.298503398895264,
-      "learning_rate": 0.0005208,
-      "loss": 2.1276,
+      "grad_norm": 2.0488903522491455,
+      "learning_rate": 0.00039029999999999995,
+      "loss": 2.2548,
       "step": 1314
     },
     {
       "epoch": 2.104,
-      "grad_norm": 1.7098180055618286,
-      "learning_rate": 0.0005212,
-      "loss": 2.061,
+      "grad_norm": 1.8532054424285889,
+      "learning_rate": 0.00039059999999999995,
+      "loss": 1.9928,
       "step": 1315
     },
     {
       "epoch": 2.1056,
-      "grad_norm": 1.733177900314331,
-      "learning_rate": 0.0005215999999999999,
-      "loss": 2.0207,
+      "grad_norm": 4.387415885925293,
+      "learning_rate": 0.00039089999999999996,
+      "loss": 2.2367,
       "step": 1316
     },
     {
       "epoch": 2.1072,
-      "grad_norm": 2.170844078063965,
-      "learning_rate": 0.000522,
-      "loss": 1.9781,
+      "grad_norm": 13.355759620666504,
+      "learning_rate": 0.00039119999999999997,
+      "loss": 3.1097,
       "step": 1317
     },
     {
       "epoch": 2.1088,
-      "grad_norm": 4.409342288970947,
-      "learning_rate": 0.0005224,
-      "loss": 2.2473,
+      "grad_norm": 6.663236141204834,
+      "learning_rate": 0.0003914999999999999,
+      "loss": 2.3162,
       "step": 1318
     },
     {
       "epoch": 2.1104,
-      "grad_norm": 3.9420828819274902,
-      "learning_rate": 0.0005227999999999999,
-      "loss": 2.2477,
+      "grad_norm": 8.847925186157227,
+      "learning_rate": 0.0003918,
+      "loss": 2.3232,
       "step": 1319
     },
     {
       "epoch": 2.112,
-      "grad_norm": 4.4460649490356445,
-      "learning_rate": 0.0005231999999999999,
-      "loss": 2.0891,
+      "grad_norm": 8.594993591308594,
+      "learning_rate": 0.00039209999999999994,
+      "loss": 2.3261,
       "step": 1320
     },
     {
       "epoch": 2.1136,
-      "grad_norm": 3.511218547821045,
-      "learning_rate": 0.0005235999999999999,
-      "loss": 2.1115,
+      "grad_norm": 7.956197261810303,
+      "learning_rate": 0.0003924,
+      "loss": 2.4478,
       "step": 1321
     },
     {
       "epoch": 2.1152,
-      "grad_norm": 4.6625142097473145,
-      "learning_rate": 0.0005239999999999999,
-      "loss": 2.0631,
+      "grad_norm": 7.233311176300049,
+      "learning_rate": 0.00039269999999999995,
+      "loss": 2.2092,
       "step": 1322
     },
     {
       "epoch": 2.1168,
-      "grad_norm": 2.5629169940948486,
-      "learning_rate": 0.0005244,
-      "loss": 2.1697,
+      "grad_norm": 4.360085487365723,
+      "learning_rate": 0.00039299999999999996,
+      "loss": 2.3606,
       "step": 1323
     },
     {
       "epoch": 2.1184,
-      "grad_norm": 0.9935573935508728,
-      "learning_rate": 0.0005248,
-      "loss": 2.055,
+      "grad_norm": 4.5778608322143555,
+      "learning_rate": 0.00039329999999999996,
+      "loss": 2.2137,
       "step": 1324
     },
     {
       "epoch": 2.12,
-      "grad_norm": 0.8370351195335388,
-      "learning_rate": 0.0005252,
-      "loss": 1.9884,
+      "grad_norm": 3.4425010681152344,
+      "learning_rate": 0.00039359999999999997,
+      "loss": 2.0536,
       "step": 1325
     },
     {
       "epoch": 2.1216,
-      "grad_norm": 1.188152551651001,
-      "learning_rate": 0.0005256,
-      "loss": 2.0258,
+      "grad_norm": 1.9287109375,
+      "learning_rate": 0.0003938999999999999,
+      "loss": 2.0203,
       "step": 1326
     },
     {
       "epoch": 2.1232,
-      "grad_norm": 1.5342425107955933,
-      "learning_rate": 0.000526,
-      "loss": 2.0623,
+      "grad_norm": 1.093762755393982,
+      "learning_rate": 0.0003942,
+      "loss": 2.1598,
       "step": 1327
     },
     {
       "epoch": 2.1248,
-      "grad_norm": 1.1506816148757935,
-      "learning_rate": 0.0005263999999999999,
-      "loss": 2.1881,
+      "grad_norm": 1.6666079759597778,
+      "learning_rate": 0.00039449999999999994,
+      "loss": 2.2012,
       "step": 1328
     },
     {
       "epoch": 2.1264,
-      "grad_norm": 1.0641168355941772,
-      "learning_rate": 0.0005267999999999999,
-      "loss": 2.0839,
+      "grad_norm": 1.5958610773086548,
+      "learning_rate": 0.0003948,
+      "loss": 1.9354,
       "step": 1329
     },
     {
       "epoch": 2.128,
-      "grad_norm": 0.6758621335029602,
-      "learning_rate": 0.0005272,
-      "loss": 2.1336,
+      "grad_norm": 1.274950385093689,
+      "learning_rate": 0.00039509999999999995,
+      "loss": 1.9601,
       "step": 1330
     },
     {
       "epoch": 2.1296,
-      "grad_norm": 2.4014053344726562,
-      "learning_rate": 0.0005275999999999999,
-      "loss": 2.1034,
+      "grad_norm": 0.9257663488388062,
+      "learning_rate": 0.00039539999999999996,
+      "loss": 2.1997,
       "step": 1331
     },
     {
       "epoch": 2.1312,
-      "grad_norm": 4.051090717315674,
-      "learning_rate": 0.0005279999999999999,
-      "loss": 2.1265,
+      "grad_norm": 0.844883918762207,
+      "learning_rate": 0.00039569999999999997,
+      "loss": 1.9768,
       "step": 1332
     },
     {
       "epoch": 2.1328,
-      "grad_norm": 4.70905876159668,
-      "learning_rate": 0.0005283999999999999,
-      "loss": 2.321,
+      "grad_norm": 3.2150275707244873,
+      "learning_rate": 0.000396,
+      "loss": 2.0972,
       "step": 1333
     },
     {
       "epoch": 2.1344,
-      "grad_norm": 2.109508752822876,
-      "learning_rate": 0.0005288,
-      "loss": 2.0784,
+      "grad_norm": 3.564847469329834,
+      "learning_rate": 0.00039629999999999993,
+      "loss": 2.1646,
       "step": 1334
     },
     {
       "epoch": 2.136,
-      "grad_norm": 3.391434907913208,
-      "learning_rate": 0.0005292,
-      "loss": 2.1631,
+      "grad_norm": 3.8981540203094482,
+      "learning_rate": 0.0003966,
+      "loss": 1.9788,
       "step": 1335
     },
     {
       "epoch": 2.1376,
-      "grad_norm": 2.689666271209717,
-      "learning_rate": 0.0005296,
-      "loss": 2.1379,
+      "grad_norm": 4.616899013519287,
+      "learning_rate": 0.00039689999999999994,
+      "loss": 2.2425,
       "step": 1336
     },
     {
       "epoch": 2.1391999999999998,
-      "grad_norm": 1.4555532932281494,
-      "learning_rate": 0.00053,
-      "loss": 2.1257,
+      "grad_norm": 4.975461483001709,
+      "learning_rate": 0.0003972,
+      "loss": 2.1961,
       "step": 1337
     },
     {
       "epoch": 2.1408,
-      "grad_norm": 1.207741379737854,
-      "learning_rate": 0.0005304,
-      "loss": 2.1133,
+      "grad_norm": 3.547156572341919,
+      "learning_rate": 0.00039749999999999996,
+      "loss": 2.1786,
       "step": 1338
     },
     {
       "epoch": 2.1424,
-      "grad_norm": 1.0733281373977661,
-      "learning_rate": 0.0005308,
-      "loss": 2.1144,
+      "grad_norm": 2.7288384437561035,
+      "learning_rate": 0.00039779999999999997,
+      "loss": 2.2044,
       "step": 1339
     },
     {
       "epoch": 2.144,
-      "grad_norm": 1.5815534591674805,
-      "learning_rate": 0.0005311999999999999,
-      "loss": 2.2798,
+      "grad_norm": 2.037722110748291,
+      "learning_rate": 0.0003981,
+      "loss": 1.9798,
       "step": 1340
     },
     {
       "epoch": 2.1456,
-      "grad_norm": 0.9271948933601379,
-      "learning_rate": 0.0005315999999999999,
-      "loss": 2.4806,
+      "grad_norm": 0.9988428354263306,
+      "learning_rate": 0.0003984,
+      "loss": 1.9068,
       "step": 1341
     },
     {
       "epoch": 2.1471999999999998,
-      "grad_norm": 1.6899449825286865,
-      "learning_rate": 0.000532,
-      "loss": 2.6045,
+      "grad_norm": 0.8600280284881592,
+      "learning_rate": 0.00039869999999999993,
+      "loss": 2.2989,
       "step": 1342
     },
     {
       "epoch": 2.1488,
-      "grad_norm": 3.4768803119659424,
-      "learning_rate": 0.0005323999999999999,
-      "loss": 2.4404,
+      "grad_norm": 0.8826497197151184,
+      "learning_rate": 0.000399,
+      "loss": 2.1012,
       "step": 1343
     },
     {
       "epoch": 2.1504,
-      "grad_norm": 3.074777603149414,
-      "learning_rate": 0.0005327999999999999,
-      "loss": 2.2315,
+      "grad_norm": 3.0037686824798584,
+      "learning_rate": 0.00039929999999999995,
+      "loss": 2.1131,
       "step": 1344
     },
     {
       "epoch": 2.152,
-      "grad_norm": 1.114790439605713,
-      "learning_rate": 0.0005332,
-      "loss": 2.2487,
+      "grad_norm": 1.0067633390426636,
+      "learning_rate": 0.0003996,
+      "loss": 2.0187,
       "step": 1345
     },
     {
       "epoch": 2.1536,
-      "grad_norm": 1.1238393783569336,
-      "learning_rate": 0.0005336,
-      "loss": 2.0376,
+      "grad_norm": 1.6712292432785034,
+      "learning_rate": 0.00039989999999999996,
+      "loss": 2.3152,
       "step": 1346
     },
     {
       "epoch": 2.1552,
-      "grad_norm": 0.6773790121078491,
-      "learning_rate": 0.000534,
-      "loss": 2.1915,
+      "grad_norm": 4.251307964324951,
+      "learning_rate": 0.00040019999999999997,
+      "loss": 1.97,
       "step": 1347
     },
     {
       "epoch": 2.1568,
-      "grad_norm": 0.7934916019439697,
-      "learning_rate": 0.0005344,
-      "loss": 2.471,
+      "grad_norm": 1.497650146484375,
+      "learning_rate": 0.0004005,
+      "loss": 2.4916,
       "step": 1348
     },
     {
       "epoch": 2.1584,
-      "grad_norm": 1.175371527671814,
-      "learning_rate": 0.0005348,
-      "loss": 2.4804,
+      "grad_norm": 1.7032194137573242,
+      "learning_rate": 0.0004008,
+      "loss": 2.2926,
       "step": 1349
     },
     {
       "epoch": 2.16,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005348,
-      "loss": 2.1641,
+      "grad_norm": 1.2858754396438599,
+      "learning_rate": 0.00040109999999999994,
+      "loss": 2.5882,
       "step": 1350
     },
     {
       "epoch": 2.1616,
-      "grad_norm": 21.404155731201172,
-      "learning_rate": 0.0005352,
-      "loss": 3.6373,
+      "grad_norm": 4.7999773025512695,
+      "learning_rate": 0.0004014,
+      "loss": 2.2346,
       "step": 1351
     },
     {
       "epoch": 2.1632,
-      "grad_norm": 3.474867105484009,
-      "learning_rate": 0.0005356,
-      "loss": 2.4769,
+      "grad_norm": 3.09006404876709,
+      "learning_rate": 0.00040169999999999995,
+      "loss": 2.2034,
       "step": 1352
     },
     {
       "epoch": 2.1648,
-      "grad_norm": 10.239166259765625,
-      "learning_rate": 0.0005359999999999999,
-      "loss": 2.9065,
+      "grad_norm": 1.7907410860061646,
+      "learning_rate": 0.000402,
+      "loss": 2.2089,
       "step": 1353
     },
     {
       "epoch": 2.1664,
-      "grad_norm": 1.5855854749679565,
-      "learning_rate": 0.0005363999999999999,
-      "loss": 2.3575,
+      "grad_norm": 3.0609290599823,
+      "learning_rate": 0.00040229999999999997,
+      "loss": 2.2091,
       "step": 1354
     },
     {
       "epoch": 2.168,
-      "grad_norm": 1.557480812072754,
-      "learning_rate": 0.0005368,
-      "loss": 2.4068,
+      "grad_norm": 3.085463285446167,
+      "learning_rate": 0.0004026,
+      "loss": 2.7211,
       "step": 1355
     },
     {
       "epoch": 2.1696,
-      "grad_norm": 3.996270179748535,
-      "learning_rate": 0.0005371999999999999,
-      "loss": 2.2568,
+      "grad_norm": 6.075860023498535,
+      "learning_rate": 0.0004029,
+      "loss": 2.3363,
       "step": 1356
     },
     {
       "epoch": 2.1712,
-      "grad_norm": 2.9636332988739014,
-      "learning_rate": 0.0005376,
-      "loss": 2.3238,
+      "grad_norm": 7.942235469818115,
+      "learning_rate": 0.0004032,
+      "loss": 2.4267,
       "step": 1357
     },
     {
       "epoch": 2.1728,
-      "grad_norm": 5.762381553649902,
-      "learning_rate": 0.000538,
-      "loss": 2.4183,
+      "grad_norm": 9.660760879516602,
+      "learning_rate": 0.00040349999999999994,
+      "loss": 2.7312,
       "step": 1358
     },
     {
       "epoch": 2.1744,
-      "grad_norm": 5.73525333404541,
-      "learning_rate": 0.0005384,
-      "loss": 2.32,
+      "grad_norm": 10.228667259216309,
+      "learning_rate": 0.0004038,
+      "loss": 2.676,
       "step": 1359
     },
     {
       "epoch": 2.176,
-      "grad_norm": 5.973430156707764,
-      "learning_rate": 0.0005388,
-      "loss": 2.2027,
+      "grad_norm": 10.17164134979248,
+      "learning_rate": 0.00040409999999999996,
+      "loss": 3.0616,
       "step": 1360
     },
     {
       "epoch": 2.1776,
-      "grad_norm": 4.767578125,
-      "learning_rate": 0.0005391999999999999,
-      "loss": 2.1613,
+      "grad_norm": 10.906268119812012,
+      "learning_rate": 0.0004044,
+      "loss": 3.132,
       "step": 1361
     },
     {
       "epoch": 2.1792,
-      "grad_norm": 4.585302352905273,
-      "learning_rate": 0.0005396,
-      "loss": 2.095,
+      "grad_norm": 10.258222579956055,
+      "learning_rate": 0.00040469999999999997,
+      "loss": 2.9752,
       "step": 1362
     },
     {
       "epoch": 2.1808,
-      "grad_norm": 2.8744828701019287,
-      "learning_rate": 0.00054,
-      "loss": 2.1879,
+      "grad_norm": 10.155888557434082,
+      "learning_rate": 0.000405,
+      "loss": 2.8489,
       "step": 1363
     },
     {
       "epoch": 2.1824,
-      "grad_norm": 2.7364563941955566,
-      "learning_rate": 0.0005403999999999999,
-      "loss": 1.8963,
+      "grad_norm": 9.648914337158203,
+      "learning_rate": 0.00040529999999999993,
+      "loss": 2.7908,
       "step": 1364
     },
     {
       "epoch": 2.184,
-      "grad_norm": 1.149092435836792,
-      "learning_rate": 0.0005407999999999999,
-      "loss": 1.7546,
+      "grad_norm": 8.77664566040039,
+      "learning_rate": 0.0004056,
+      "loss": 2.5796,
       "step": 1365
     },
     {
       "epoch": 2.1856,
-      "grad_norm": 1.0585405826568604,
-      "learning_rate": 0.0005411999999999999,
-      "loss": 1.8459,
+      "grad_norm": 7.69537878036499,
+      "learning_rate": 0.00040589999999999995,
+      "loss": 2.5244,
       "step": 1366
     },
     {
       "epoch": 2.1872,
-      "grad_norm": 1.0302655696868896,
-      "learning_rate": 0.0005415999999999999,
-      "loss": 1.9024,
+      "grad_norm": 6.346115589141846,
+      "learning_rate": 0.0004062,
+      "loss": 2.409,
       "step": 1367
     },
     {
       "epoch": 2.1888,
-      "grad_norm": 4.1598801612854,
-      "learning_rate": 0.000542,
-      "loss": 2.2494,
+      "grad_norm": 4.901920795440674,
+      "learning_rate": 0.00040649999999999996,
+      "loss": 2.3376,
       "step": 1368
     },
     {
       "epoch": 2.1904,
-      "grad_norm": 3.350557565689087,
-      "learning_rate": 0.0005424,
-      "loss": 1.8213,
+      "grad_norm": 2.19638991355896,
+      "learning_rate": 0.00040679999999999997,
+      "loss": 2.1,
       "step": 1369
     },
     {
       "epoch": 2.192,
-      "grad_norm": 5.341938495635986,
-      "learning_rate": 0.0005428,
-      "loss": 2.008,
+      "grad_norm": 1.7705507278442383,
+      "learning_rate": 0.0004071,
+      "loss": 2.0965,
       "step": 1370
     },
     {
       "epoch": 2.1936,
-      "grad_norm": 3.1162655353546143,
-      "learning_rate": 0.0005432,
-      "loss": 2.006,
+      "grad_norm": 5.908679485321045,
+      "learning_rate": 0.0004074,
+      "loss": 2.3494,
       "step": 1371
     },
     {
       "epoch": 2.1952,
-      "grad_norm": 3.5952706336975098,
-      "learning_rate": 0.0005436,
-      "loss": 2.0343,
+      "grad_norm": 4.028108596801758,
+      "learning_rate": 0.00040769999999999994,
+      "loss": 2.3946,
       "step": 1372
     },
     {
       "epoch": 2.1968,
-      "grad_norm": 1.513656497001648,
-      "learning_rate": 0.0005439999999999999,
-      "loss": 2.2091,
+      "grad_norm": 6.116652488708496,
+      "learning_rate": 0.000408,
+      "loss": 2.223,
       "step": 1373
     },
     {
       "epoch": 2.1984,
-      "grad_norm": 1.1791319847106934,
-      "learning_rate": 0.0005443999999999999,
-      "loss": 2.1118,
+      "grad_norm": 7.863028526306152,
+      "learning_rate": 0.00040829999999999995,
+      "loss": 2.6158,
       "step": 1374
     },
     {
       "epoch": 2.2,
-      "grad_norm": 1.423629879951477,
-      "learning_rate": 0.0005448,
-      "loss": 1.9229,
+      "grad_norm": 7.65760612487793,
+      "learning_rate": 0.0004086,
+      "loss": 2.2629,
       "step": 1375
     },
     {
       "epoch": 2.2016,
-      "grad_norm": 1.9064466953277588,
-      "learning_rate": 0.0005451999999999999,
-      "loss": 1.8431,
+      "grad_norm": 6.600236415863037,
+      "learning_rate": 0.00040889999999999996,
+      "loss": 2.1404,
       "step": 1376
     },
     {
       "epoch": 2.2032,
-      "grad_norm": 1.9968523979187012,
-      "learning_rate": 0.0005455999999999999,
-      "loss": 1.8884,
+      "grad_norm": 7.769944190979004,
+      "learning_rate": 0.00040919999999999997,
+      "loss": 2.6362,
       "step": 1377
     },
     {
       "epoch": 2.2048,
-      "grad_norm": 2.7741806507110596,
-      "learning_rate": 0.0005459999999999999,
-      "loss": 2.0129,
+      "grad_norm": 6.450319766998291,
+      "learning_rate": 0.0004095,
+      "loss": 2.3437,
       "step": 1378
     },
     {
       "epoch": 2.2064,
-      "grad_norm": 2.164896011352539,
-      "learning_rate": 0.0005463999999999999,
-      "loss": 1.8725,
+      "grad_norm": 6.125224590301514,
+      "learning_rate": 0.0004098,
+      "loss": 2.2856,
       "step": 1379
     },
     {
       "epoch": 2.208,
-      "grad_norm": 2.3117516040802,
-      "learning_rate": 0.0005468,
-      "loss": 2.1264,
+      "grad_norm": 3.7598607540130615,
+      "learning_rate": 0.00041009999999999994,
+      "loss": 2.4335,
       "step": 1380
     },
     {
       "epoch": 2.2096,
-      "grad_norm": 1.8635528087615967,
-      "learning_rate": 0.0005472,
-      "loss": 1.8341,
+      "grad_norm": 2.4773128032684326,
+      "learning_rate": 0.0004104,
+      "loss": 2.1278,
       "step": 1381
     },
     {
       "epoch": 2.2112,
-      "grad_norm": 0.991112470626831,
-      "learning_rate": 0.0005476,
-      "loss": 2.1591,
+      "grad_norm": 1.4773629903793335,
+      "learning_rate": 0.00041069999999999995,
+      "loss": 2.2617,
       "step": 1382
     },
     {
       "epoch": 2.2128,
-      "grad_norm": 1.4743876457214355,
-      "learning_rate": 0.000548,
-      "loss": 2.1582,
+      "grad_norm": 1.111156702041626,
+      "learning_rate": 0.000411,
+      "loss": 2.2365,
       "step": 1383
     },
     {
       "epoch": 2.2144,
-      "grad_norm": 0.8621456027030945,
-      "learning_rate": 0.0005484,
-      "loss": 1.7388,
+      "grad_norm": 1.0154348611831665,
+      "learning_rate": 0.00041129999999999997,
+      "loss": 2.2948,
       "step": 1384
     },
     {
       "epoch": 2.216,
-      "grad_norm": 0.8471877574920654,
-      "learning_rate": 0.0005487999999999999,
-      "loss": 1.9254,
+      "grad_norm": 0.8806530833244324,
+      "learning_rate": 0.0004116,
+      "loss": 1.9224,
       "step": 1385
     },
     {
       "epoch": 2.2176,
-      "grad_norm": 1.153436303138733,
-      "learning_rate": 0.0005491999999999999,
-      "loss": 1.9162,
+      "grad_norm": 1.5271722078323364,
+      "learning_rate": 0.0004119,
+      "loss": 2.0459,
       "step": 1386
     },
     {
       "epoch": 2.2192,
-      "grad_norm": 1.1189401149749756,
-      "learning_rate": 0.0005496,
-      "loss": 2.1507,
+      "grad_norm": 1.2786515951156616,
+      "learning_rate": 0.0004122,
+      "loss": 1.9195,
       "step": 1387
     },
     {
       "epoch": 2.2208,
-      "grad_norm": 0.7863136529922485,
-      "learning_rate": 0.0005499999999999999,
-      "loss": 1.917,
+      "grad_norm": 2.0746285915374756,
+      "learning_rate": 0.00041249999999999994,
+      "loss": 2.1174,
       "step": 1388
     },
     {
       "epoch": 2.2224,
-      "grad_norm": 0.6574200391769409,
-      "learning_rate": 0.0005503999999999999,
-      "loss": 2.1708,
+      "grad_norm": 1.9179648160934448,
+      "learning_rate": 0.00041279999999999995,
+      "loss": 2.0782,
       "step": 1389
     },
     {
       "epoch": 2.224,
-      "grad_norm": 1.5995211601257324,
-      "learning_rate": 0.0005507999999999999,
-      "loss": 1.8246,
+      "grad_norm": 3.8180932998657227,
+      "learning_rate": 0.00041309999999999996,
+      "loss": 2.2007,
       "step": 1390
     },
     {
       "epoch": 2.2256,
-      "grad_norm": 1.0908114910125732,
-      "learning_rate": 0.0005512,
-      "loss": 2.2587,
+      "grad_norm": 2.905001163482666,
+      "learning_rate": 0.0004133999999999999,
+      "loss": 2.4089,
       "step": 1391
     },
     {
       "epoch": 2.2272,
-      "grad_norm": 0.8126534223556519,
-      "learning_rate": 0.0005516,
-      "loss": 2.077,
+      "grad_norm": 1.348793864250183,
+      "learning_rate": 0.00041369999999999997,
+      "loss": 2.3518,
       "step": 1392
     },
     {
       "epoch": 2.2288,
-      "grad_norm": 0.8407979607582092,
-      "learning_rate": 0.000552,
-      "loss": 2.0135,
+      "grad_norm": 0.6599010229110718,
+      "learning_rate": 0.0004139999999999999,
+      "loss": 2.0014,
       "step": 1393
     },
     {
       "epoch": 2.2304,
-      "grad_norm": 8.15050983428955,
-      "learning_rate": 0.0005524,
-      "loss": 2.8744,
+      "grad_norm": 0.8205379247665405,
+      "learning_rate": 0.0004143,
+      "loss": 2.0897,
       "step": 1394
     },
     {
       "epoch": 2.232,
-      "grad_norm": 1.2660044431686401,
-      "learning_rate": 0.0005528,
-      "loss": 2.0023,
+      "grad_norm": 1.9124035835266113,
+      "learning_rate": 0.00041459999999999994,
+      "loss": 2.2603,
       "step": 1395
     },
     {
       "epoch": 2.2336,
-      "grad_norm": 0.7754627466201782,
-      "learning_rate": 0.0005532,
-      "loss": 2.2246,
+      "grad_norm": 0.7764268517494202,
+      "learning_rate": 0.00041489999999999995,
+      "loss": 2.1528,
       "step": 1396
     },
     {
       "epoch": 2.2352,
-      "grad_norm": 1.0184060335159302,
-      "learning_rate": 0.0005535999999999999,
-      "loss": 2.4245,
+      "grad_norm": 0.5789710879325867,
+      "learning_rate": 0.00041519999999999995,
+      "loss": 2.1246,
       "step": 1397
     },
     {
       "epoch": 2.2368,
-      "grad_norm": 0.7693429589271545,
-      "learning_rate": 0.0005539999999999999,
-      "loss": 2.5069,
+      "grad_norm": 1.1585594415664673,
+      "learning_rate": 0.00041549999999999996,
+      "loss": 2.224,
       "step": 1398
     },
     {
       "epoch": 2.2384,
-      "grad_norm": 1.50884211063385,
-      "learning_rate": 0.0005544,
-      "loss": 2.4099,
+      "grad_norm": 0.8179572820663452,
+      "learning_rate": 0.0004157999999999999,
+      "loss": 2.5131,
       "step": 1399
     },
     {
       "epoch": 2.24,
-      "grad_norm": 1.125415563583374,
-      "learning_rate": 0.0005547999999999999,
-      "loss": 2.6928,
-      "step": 1400
-    },
-    {
-      "epoch": 2.24,
-      "eval_cer": 0.49043727827161476,
-      "eval_loss": 2.3595921993255615,
-      "eval_runtime": 159.1952,
-      "eval_samples_per_second": 19.699,
-      "eval_steps_per_second": 1.231,
-      "eval_wer": 0.7915449176251166,
+      "grad_norm": 1.2048473358154297,
+      "learning_rate": 0.0004161,
+      "loss": 2.7738,
       "step": 1400
     },
     {
       "epoch": 2.2416,
-      "grad_norm": 9.09476375579834,
-      "learning_rate": 0.0005551999999999999,
-      "loss": 2.6805,
+      "grad_norm": 8.999752044677734,
+      "learning_rate": 0.00041639999999999993,
+      "loss": 2.6463,
       "step": 1401
     },
     {
       "epoch": 2.2432,
-      "grad_norm": 5.154038906097412,
-      "learning_rate": 0.0005556,
-      "loss": 2.4796,
+      "grad_norm": 2.3112637996673584,
+      "learning_rate": 0.0004167,
+      "loss": 2.2288,
       "step": 1402
     },
     {
       "epoch": 2.2448,
-      "grad_norm": 11.566636085510254,
-      "learning_rate": 0.000556,
-      "loss": 2.8707,
+      "grad_norm": 4.989300727844238,
+      "learning_rate": 0.00041699999999999994,
+      "loss": 2.4311,
       "step": 1403
     },
     {
       "epoch": 2.2464,
-      "grad_norm": 6.700075626373291,
-      "learning_rate": 0.0005564,
-      "loss": 2.5268,
+      "grad_norm": 1.8407671451568604,
+      "learning_rate": 0.00041729999999999995,
+      "loss": 2.3069,
       "step": 1404
     },
     {
       "epoch": 2.248,
-      "grad_norm": 3.7919178009033203,
-      "learning_rate": 0.0005568,
-      "loss": 2.3305,
+      "grad_norm": 1.7011024951934814,
+      "learning_rate": 0.00041759999999999996,
+      "loss": 2.4687,
       "step": 1405
     },
     {
       "epoch": 2.2496,
-      "grad_norm": 0.8292596936225891,
-      "learning_rate": 0.0005571999999999999,
-      "loss": 2.1945,
+      "grad_norm": 1.005672574043274,
+      "learning_rate": 0.00041789999999999997,
+      "loss": 2.416,
       "step": 1406
     },
     {
       "epoch": 2.2512,
-      "grad_norm": 0.7672077417373657,
-      "learning_rate": 0.0005576,
-      "loss": 2.2364,
+      "grad_norm": 4.635334014892578,
+      "learning_rate": 0.0004181999999999999,
+      "loss": 2.2739,
       "step": 1407
     },
     {
       "epoch": 2.2528,
-      "grad_norm": 2.8418593406677246,
-      "learning_rate": 0.000558,
-      "loss": 2.1488,
+      "grad_norm": 5.972519397735596,
+      "learning_rate": 0.0004185,
+      "loss": 2.2631,
       "step": 1408
     },
     {
       "epoch": 2.2544,
-      "grad_norm": 1.9393724203109741,
-      "learning_rate": 0.0005583999999999999,
-      "loss": 2.3457,
+      "grad_norm": 6.264317512512207,
+      "learning_rate": 0.00041879999999999993,
+      "loss": 2.177,
       "step": 1409
     },
     {
       "epoch": 2.2560000000000002,
-      "grad_norm": 3.298280715942383,
-      "learning_rate": 0.0005587999999999999,
-      "loss": 2.2249,
+      "grad_norm": 6.628238677978516,
+      "learning_rate": 0.0004191,
+      "loss": 2.289,
       "step": 1410
     },
     {
       "epoch": 2.2576,
-      "grad_norm": 4.188957214355469,
-      "learning_rate": 0.0005591999999999999,
-      "loss": 2.1751,
+      "grad_norm": 5.494356155395508,
+      "learning_rate": 0.00041939999999999995,
+      "loss": 2.3197,
       "step": 1411
     },
     {
       "epoch": 2.2592,
-      "grad_norm": 1.6427080631256104,
-      "learning_rate": 0.0005595999999999999,
-      "loss": 2.2901,
+      "grad_norm": 2.957184314727783,
+      "learning_rate": 0.00041969999999999996,
+      "loss": 2.1555,
       "step": 1412
     },
     {
       "epoch": 2.2608,
-      "grad_norm": 1.5279046297073364,
-      "learning_rate": 0.00056,
-      "loss": 2.0703,
+      "grad_norm": 3.3862264156341553,
+      "learning_rate": 0.00041999999999999996,
+      "loss": 2.1691,
       "step": 1413
     },
     {
       "epoch": 2.2624,
-      "grad_norm": 1.4491573572158813,
-      "learning_rate": 0.0005604,
-      "loss": 2.1491,
+      "grad_norm": 2.009979486465454,
+      "learning_rate": 0.00042029999999999997,
+      "loss": 1.9896,
       "step": 1414
     },
     {
       "epoch": 2.2640000000000002,
-      "grad_norm": 0.9883561730384827,
-      "learning_rate": 0.0005608,
-      "loss": 2.1274,
+      "grad_norm": 0.8138348460197449,
+      "learning_rate": 0.0004205999999999999,
+      "loss": 2.0016,
       "step": 1415
     },
     {
       "epoch": 2.2656,
-      "grad_norm": 5.221831321716309,
-      "learning_rate": 0.0005612,
-      "loss": 2.2393,
+      "grad_norm": 1.9805331230163574,
+      "learning_rate": 0.0004209,
+      "loss": 2.0405,
       "step": 1416
     },
     {
       "epoch": 2.2672,
-      "grad_norm": 3.9041683673858643,
-      "learning_rate": 0.0005616,
-      "loss": 2.2982,
+      "grad_norm": 3.7756824493408203,
+      "learning_rate": 0.00042119999999999994,
+      "loss": 2.2341,
       "step": 1417
     },
     {
       "epoch": 2.2688,
-      "grad_norm": 3.2574350833892822,
-      "learning_rate": 0.0005619999999999999,
-      "loss": 2.3462,
+      "grad_norm": 4.360873222351074,
+      "learning_rate": 0.00042149999999999995,
+      "loss": 1.9849,
       "step": 1418
     },
     {
       "epoch": 2.2704,
-      "grad_norm": 2.5626466274261475,
-      "learning_rate": 0.0005624,
-      "loss": 2.3557,
+      "grad_norm": 4.901808738708496,
+      "learning_rate": 0.00042179999999999995,
+      "loss": 2.1964,
       "step": 1419
     },
     {
       "epoch": 2.2720000000000002,
-      "grad_norm": 6.10444974899292,
-      "learning_rate": 0.0005627999999999999,
-      "loss": 2.3624,
+      "grad_norm": 5.945872783660889,
+      "learning_rate": 0.00042209999999999996,
+      "loss": 2.2691,
       "step": 1420
     },
     {
       "epoch": 2.2736,
-      "grad_norm": 7.629668712615967,
-      "learning_rate": 0.0005631999999999999,
-      "loss": 2.5705,
+      "grad_norm": 6.475261211395264,
+      "learning_rate": 0.0004223999999999999,
+      "loss": 2.3456,
       "step": 1421
     },
     {
       "epoch": 2.2752,
-      "grad_norm": 0.9113142490386963,
-      "learning_rate": 0.0005635999999999999,
-      "loss": 2.2048,
+      "grad_norm": 5.20826530456543,
+      "learning_rate": 0.0004227,
+      "loss": 2.3354,
       "step": 1422
     },
     {
       "epoch": 2.2768,
-      "grad_norm": 2.0822792053222656,
-      "learning_rate": 0.0005639999999999999,
-      "loss": 2.2642,
+      "grad_norm": 5.351243019104004,
+      "learning_rate": 0.00042299999999999993,
+      "loss": 2.4577,
       "step": 1423
     },
     {
       "epoch": 2.2784,
-      "grad_norm": 3.2671005725860596,
-      "learning_rate": 0.0005644,
-      "loss": 2.3544,
+      "grad_norm": 3.6445577144622803,
+      "learning_rate": 0.0004233,
+      "loss": 2.2532,
       "step": 1424
     },
     {
       "epoch": 2.2800000000000002,
-      "grad_norm": 0.9781720042228699,
-      "learning_rate": 0.0005648,
-      "loss": 2.5447,
+      "grad_norm": 2.2167351245880127,
+      "learning_rate": 0.00042359999999999994,
+      "loss": 2.2576,
       "step": 1425
     },
     {
       "epoch": 2.2816,
-      "grad_norm": 3.4515535831451416,
-      "learning_rate": 0.0005652,
-      "loss": 2.4399,
+      "grad_norm": 1.7759617567062378,
+      "learning_rate": 0.00042389999999999995,
+      "loss": 2.3795,
       "step": 1426
     },
     {
       "epoch": 2.2832,
-      "grad_norm": 6.058647155761719,
-      "learning_rate": 0.0005656,
-      "loss": 2.3394,
+      "grad_norm": 1.0135295391082764,
+      "learning_rate": 0.00042419999999999996,
+      "loss": 2.3208,
       "step": 1427
     },
     {
       "epoch": 2.2848,
-      "grad_norm": 5.424431324005127,
-      "learning_rate": 0.000566,
-      "loss": 2.3398,
+      "grad_norm": 2.575662851333618,
+      "learning_rate": 0.00042449999999999996,
+      "loss": 2.1992,
       "step": 1428
     },
     {
       "epoch": 2.2864,
-      "grad_norm": 5.348411560058594,
-      "learning_rate": 0.0005663999999999999,
-      "loss": 2.4424,
+      "grad_norm": 0.6912661194801331,
+      "learning_rate": 0.0004247999999999999,
+      "loss": 2.2395,
       "step": 1429
     },
     {
       "epoch": 2.288,
-      "grad_norm": 2.7252535820007324,
-      "learning_rate": 0.0005667999999999999,
-      "loss": 2.1842,
+      "grad_norm": 0.6977083086967468,
+      "learning_rate": 0.0004251,
+      "loss": 2.079,
       "step": 1430
     },
     {
       "epoch": 2.2896,
-      "grad_norm": 1.38560950756073,
-      "learning_rate": 0.0005672,
-      "loss": 2.2131,
+      "grad_norm": 1.251266598701477,
+      "learning_rate": 0.00042539999999999993,
+      "loss": 2.2202,
       "step": 1431
     },
     {
       "epoch": 2.2912,
-      "grad_norm": 0.8164823055267334,
-      "learning_rate": 0.0005675999999999999,
-      "loss": 2.2275,
+      "grad_norm": 8.140886306762695,
+      "learning_rate": 0.0004257,
+      "loss": 2.5216,
       "step": 1432
     },
     {
       "epoch": 2.2928,
-      "grad_norm": 1.8317232131958008,
-      "learning_rate": 0.0005679999999999999,
-      "loss": 2.0737,
+      "grad_norm": 2.424851894378662,
+      "learning_rate": 0.00042599999999999995,
+      "loss": 2.1489,
       "step": 1433
     },
     {
       "epoch": 2.2944,
-      "grad_norm": 4.063267707824707,
-      "learning_rate": 0.0005683999999999999,
-      "loss": 2.4934,
+      "grad_norm": 1.4449694156646729,
+      "learning_rate": 0.00042629999999999995,
+      "loss": 2.1204,
       "step": 1434
     },
     {
       "epoch": 2.296,
-      "grad_norm": 2.9908406734466553,
-      "learning_rate": 0.0005688,
-      "loss": 2.3033,
+      "grad_norm": 0.626359224319458,
+      "learning_rate": 0.00042659999999999996,
+      "loss": 2.1644,
       "step": 1435
     },
     {
       "epoch": 2.2976,
-      "grad_norm": 4.536923885345459,
-      "learning_rate": 0.0005692,
-      "loss": 2.4368,
+      "grad_norm": 0.8110465407371521,
+      "learning_rate": 0.00042689999999999997,
+      "loss": 2.2867,
       "step": 1436
     },
     {
       "epoch": 2.2992,
-      "grad_norm": 3.1720566749572754,
-      "learning_rate": 0.0005696,
-      "loss": 2.246,
+      "grad_norm": 0.7100591659545898,
+      "learning_rate": 0.0004271999999999999,
+      "loss": 2.057,
       "step": 1437
     },
     {
       "epoch": 2.3008,
-      "grad_norm": 3.758767604827881,
-      "learning_rate": 0.00057,
-      "loss": 2.5514,
+      "grad_norm": 1.4392229318618774,
+      "learning_rate": 0.0004275,
+      "loss": 2.1222,
       "step": 1438
     },
     {
       "epoch": 2.3024,
-      "grad_norm": 0.8983809947967529,
-      "learning_rate": 0.0005704,
-      "loss": 2.3606,
+      "grad_norm": 4.126809120178223,
+      "learning_rate": 0.00042779999999999994,
+      "loss": 2.4204,
       "step": 1439
     },
     {
       "epoch": 2.304,
-      "grad_norm": 1.1368706226348877,
-      "learning_rate": 0.0005708,
-      "loss": 2.276,
+      "grad_norm": 2.7679941654205322,
+      "learning_rate": 0.0004281,
+      "loss": 2.413,
       "step": 1440
     },
     {
       "epoch": 2.3056,
-      "grad_norm": 0.8187576532363892,
-      "learning_rate": 0.0005711999999999999,
-      "loss": 2.3264,
+      "grad_norm": 0.9221087694168091,
+      "learning_rate": 0.00042839999999999995,
+      "loss": 2.3055,
       "step": 1441
     },
     {
       "epoch": 2.3072,
-      "grad_norm": 1.600205421447754,
-      "learning_rate": 0.0005715999999999999,
-      "loss": 2.4888,
+      "grad_norm": 1.424880027770996,
+      "learning_rate": 0.00042869999999999996,
+      "loss": 2.2364,
       "step": 1442
     },
     {
       "epoch": 2.3088,
-      "grad_norm": 1.1413285732269287,
-      "learning_rate": 0.0005719999999999999,
-      "loss": 2.2784,
+      "grad_norm": 1.1211133003234863,
+      "learning_rate": 0.00042899999999999997,
+      "loss": 2.0543,
       "step": 1443
     },
     {
       "epoch": 2.3104,
-      "grad_norm": 0.8480069637298584,
-      "learning_rate": 0.0005723999999999999,
-      "loss": 2.4388,
+      "grad_norm": 1.0176845788955688,
+      "learning_rate": 0.00042929999999999997,
+      "loss": 1.9312,
       "step": 1444
     },
     {
       "epoch": 2.312,
-      "grad_norm": 2.6615993976593018,
-      "learning_rate": 0.0005727999999999999,
-      "loss": 2.2917,
+      "grad_norm": 1.0180336236953735,
+      "learning_rate": 0.0004295999999999999,
+      "loss": 2.1439,
       "step": 1445
     },
     {
       "epoch": 2.3136,
-      "grad_norm": 2.1041688919067383,
-      "learning_rate": 0.0005732,
-      "loss": 2.432,
+      "grad_norm": 2.1806435585021973,
+      "learning_rate": 0.0004299,
+      "loss": 2.311,
       "step": 1446
     },
     {
       "epoch": 2.3152,
-      "grad_norm": 3.780215263366699,
-      "learning_rate": 0.0005736,
-      "loss": 2.5987,
+      "grad_norm": 1.58607816696167,
+      "learning_rate": 0.00043019999999999994,
+      "loss": 2.2012,
       "step": 1447
     },
     {
       "epoch": 2.3168,
-      "grad_norm": 1.8719847202301025,
-      "learning_rate": 0.000574,
-      "loss": 2.6563,
+      "grad_norm": 3.318199634552002,
+      "learning_rate": 0.0004305,
+      "loss": 2.4078,
       "step": 1448
     },
     {
       "epoch": 2.3184,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 2.4196,
+      "grad_norm": 2.1914138793945312,
+      "learning_rate": 0.00043079999999999995,
+      "loss": 2.4591,
       "step": 1449
     },
     {
       "epoch": 2.32,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 2.709254264831543,
+      "learning_rate": 0.00043109999999999996,
+      "loss": 2.4848,
       "step": 1450
     },
     {
       "epoch": 2.3216,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 4.738913536071777,
+      "learning_rate": 0.00043139999999999997,
+      "loss": 2.6373,
       "step": 1451
     },
     {
       "epoch": 2.3232,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 3.7292051315307617,
+      "learning_rate": 0.0004317,
+      "loss": 2.5686,
       "step": 1452
     },
     {
       "epoch": 2.3247999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 1.3130608797073364,
+      "learning_rate": 0.00043199999999999993,
+      "loss": 2.4265,
       "step": 1453
     },
     {
       "epoch": 2.3264,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 1.3804,
+      "grad_norm": 6.308412551879883,
+      "learning_rate": 0.0004323,
+      "loss": 2.4777,
       "step": 1454
     },
     {
       "epoch": 2.328,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 5.950598239898682,
+      "learning_rate": 0.00043259999999999994,
+      "loss": 2.3441,
       "step": 1455
     },
     {
       "epoch": 2.3296,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 7.419227600097656,
+      "learning_rate": 0.0004329,
+      "loss": 2.3924,
       "step": 1456
     },
     {
       "epoch": 2.3312,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 8.05959415435791,
+      "learning_rate": 0.00043319999999999996,
+      "loss": 2.4859,
       "step": 1457
     },
     {
       "epoch": 2.3327999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 7.334360122680664,
+      "learning_rate": 0.00043349999999999997,
+      "loss": 2.233,
       "step": 1458
     },
     {
       "epoch": 2.3344,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 6.437933444976807,
+      "learning_rate": 0.0004338,
+      "loss": 2.2507,
       "step": 1459
     },
     {
       "epoch": 2.336,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 6.478224754333496,
+      "learning_rate": 0.0004341,
+      "loss": 2.2895,
       "step": 1460
     },
     {
       "epoch": 2.3376,
-      "grad_norm": NaN,
-      "learning_rate": 0.000574,
-      "loss": 0.0,
+      "grad_norm": 5.727662563323975,
+      "learning_rate": 0.00043439999999999993,
+      "loss": 2.0447,
       "step": 1461
     },
     {
       "epoch": 2.3392,
-      "grad_norm": 1.6323294639587402,
-      "learning_rate": 0.0005744,
-      "loss": 2.3085,
+      "grad_norm": 3.6567492485046387,
+      "learning_rate": 0.0004347,
+      "loss": 2.0184,
       "step": 1462
     },
     {
       "epoch": 2.3407999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.0759,
+      "grad_norm": 1.113254427909851,
+      "learning_rate": 0.00043499999999999995,
+      "loss": 1.9926,
       "step": 1463
     },
     {
       "epoch": 2.3424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 0.7796720862388611,
+      "learning_rate": 0.00043529999999999996,
+      "loss": 2.1872,
       "step": 1464
     },
     {
       "epoch": 2.344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.1082,
+      "grad_norm": 0.9296607375144958,
+      "learning_rate": 0.00043559999999999996,
+      "loss": 2.1299,
       "step": 1465
     },
     {
       "epoch": 2.3456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.0827059745788574,
+      "learning_rate": 0.00043589999999999997,
+      "loss": 1.88,
       "step": 1466
     },
     {
       "epoch": 2.3472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.31209659576416,
+      "learning_rate": 0.0004361999999999999,
+      "loss": 1.9073,
       "step": 1467
     },
     {
       "epoch": 2.3487999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 4.975152492523193,
+      "learning_rate": 0.0004365,
+      "loss": 2.2116,
       "step": 1468
     },
     {
       "epoch": 2.3504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 6.011470794677734,
+      "learning_rate": 0.00043679999999999994,
+      "loss": 1.8669,
       "step": 1469
     },
     {
       "epoch": 2.352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.388676404953003,
+      "learning_rate": 0.0004371,
+      "loss": 2.0661,
       "step": 1470
     },
     {
       "epoch": 2.3536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.6890642642974854,
+      "learning_rate": 0.00043739999999999995,
+      "loss": 1.9085,
       "step": 1471
     },
     {
       "epoch": 2.3552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.4474709033966064,
+      "learning_rate": 0.00043769999999999996,
+      "loss": 1.9495,
       "step": 1472
     },
     {
       "epoch": 2.3568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.906247854232788,
+      "learning_rate": 0.00043799999999999997,
+      "loss": 2.3595,
       "step": 1473
     },
     {
       "epoch": 2.3584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.7890042066574097,
+      "learning_rate": 0.0004383,
+      "loss": 2.0283,
       "step": 1474
     },
     {
       "epoch": 2.36,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.4251508712768555,
+      "learning_rate": 0.00043859999999999993,
+      "loss": 2.3857,
       "step": 1475
     },
     {
       "epoch": 2.3616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.0929760932922363,
+      "learning_rate": 0.0004389,
+      "loss": 1.8558,
       "step": 1476
     },
     {
       "epoch": 2.3632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.920934796333313,
+      "learning_rate": 0.00043919999999999994,
+      "loss": 2.1394,
       "step": 1477
     },
     {
       "epoch": 2.3648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.799728274345398,
+      "learning_rate": 0.0004395,
+      "loss": 2.033,
       "step": 1478
     },
     {
       "epoch": 2.3664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.0893006324768066,
+      "learning_rate": 0.00043979999999999996,
+      "loss": 2.0484,
       "step": 1479
     },
     {
       "epoch": 2.368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.1866,
+      "grad_norm": 0.7367232441902161,
+      "learning_rate": 0.00044009999999999996,
+      "loss": 2.1303,
       "step": 1480
     },
     {
       "epoch": 2.3696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.518019437789917,
+      "learning_rate": 0.00044039999999999997,
+      "loss": 2.1944,
       "step": 1481
     },
     {
       "epoch": 2.3712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 4.497959136962891,
+      "learning_rate": 0.0004407,
+      "loss": 2.1217,
       "step": 1482
     },
     {
       "epoch": 2.3728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.2496092319488525,
+      "learning_rate": 0.00044099999999999993,
+      "loss": 2.0976,
       "step": 1483
     },
     {
       "epoch": 2.3744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.0006887912750244,
+      "learning_rate": 0.0004413,
+      "loss": 2.062,
       "step": 1484
     },
     {
       "epoch": 2.376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.2168,
+      "grad_norm": 2.9096486568450928,
+      "learning_rate": 0.00044159999999999995,
+      "loss": 2.2014,
       "step": 1485
     },
     {
       "epoch": 2.3776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.12,
+      "grad_norm": 3.6192102432250977,
+      "learning_rate": 0.0004419,
+      "loss": 2.0639,
       "step": 1486
     },
     {
       "epoch": 2.3792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.4094703197479248,
+      "learning_rate": 0.00044219999999999996,
+      "loss": 2.1488,
       "step": 1487
     },
     {
       "epoch": 2.3808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.171118140220642,
+      "learning_rate": 0.00044249999999999997,
+      "loss": 2.0222,
       "step": 1488
     },
     {
       "epoch": 2.3824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.1949,
+      "grad_norm": 1.4760013818740845,
+      "learning_rate": 0.0004428,
+      "loss": 2.0528,
       "step": 1489
     },
     {
       "epoch": 2.384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.7218807935714722,
+      "learning_rate": 0.0004431,
+      "loss": 2.0762,
       "step": 1490
     },
     {
       "epoch": 2.3856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.5412,
+      "grad_norm": 0.8926534056663513,
+      "learning_rate": 0.00044339999999999994,
+      "loss": 2.3282,
       "step": 1491
     },
     {
       "epoch": 2.3872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.467978596687317,
+      "learning_rate": 0.0004437,
+      "loss": 2.1324,
       "step": 1492
     },
     {
       "epoch": 2.3888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.302968978881836,
+      "learning_rate": 0.00044399999999999995,
+      "loss": 2.2981,
       "step": 1493
     },
     {
       "epoch": 2.3904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.0114901065826416,
+      "learning_rate": 0.0004443,
+      "loss": 2.2011,
       "step": 1494
     },
     {
       "epoch": 2.392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 4.253175258636475,
+      "learning_rate": 0.00044459999999999996,
+      "loss": 2.1162,
       "step": 1495
     },
     {
       "epoch": 2.3936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.1896846294403076,
+      "learning_rate": 0.00044489999999999997,
+      "loss": 2.2125,
       "step": 1496
     },
     {
       "epoch": 2.3952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.2957,
+      "grad_norm": 1.5464539527893066,
+      "learning_rate": 0.0004452,
+      "loss": 2.602,
       "step": 1497
     },
     {
       "epoch": 2.3968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.4599,
+      "grad_norm": 1.3925230503082275,
+      "learning_rate": 0.0004455,
+      "loss": 2.6023,
       "step": 1498
     },
     {
       "epoch": 2.3984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.246,
+      "grad_norm": 2.2040951251983643,
+      "learning_rate": 0.00044579999999999994,
+      "loss": 2.4529,
       "step": 1499
     },
     {
       "epoch": 2.4,
       "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
-      "step": 1500
-    },
-    {
-      "epoch": 2.4,
-      "eval_cer": 0.6018615219037748,
-      "eval_loss": 2.4470601081848145,
-      "eval_runtime": 158.9611,
-      "eval_samples_per_second": 19.728,
-      "eval_steps_per_second": 1.233,
-      "eval_wer": 0.8782302352087866,
+      "learning_rate": 0.00044579999999999994,
+      "loss": 1.9781,
       "step": 1500
     },
     {
       "epoch": 2.4016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.5158798694610596,
+      "learning_rate": 0.0004461,
+      "loss": 2.7577,
       "step": 1501
     },
     {
       "epoch": 2.4032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 0.7932011485099792,
+      "learning_rate": 0.00044639999999999995,
+      "loss": 2.4814,
       "step": 1502
     },
     {
       "epoch": 2.4048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.7924387454986572,
+      "learning_rate": 0.0004467,
+      "loss": 2.3667,
       "step": 1503
     },
     {
       "epoch": 2.4064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 4.5760626792907715,
+      "learning_rate": 0.00044699999999999997,
+      "loss": 2.3983,
       "step": 1504
     },
     {
       "epoch": 2.408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.0172371864318848,
+      "learning_rate": 0.0004473,
+      "loss": 2.4891,
       "step": 1505
     },
     {
       "epoch": 2.4096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 5.048511028289795,
+      "learning_rate": 0.0004476,
+      "loss": 2.5052,
       "step": 1506
     },
     {
       "epoch": 2.4112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 4.192890167236328,
+      "learning_rate": 0.0004479,
+      "loss": 2.3123,
       "step": 1507
     },
     {
       "epoch": 2.4128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.8736287355422974,
+      "learning_rate": 0.00044819999999999994,
+      "loss": 2.4851,
       "step": 1508
     },
     {
       "epoch": 2.4144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.5078847408294678,
+      "learning_rate": 0.0004485,
+      "loss": 2.3538,
       "step": 1509
     },
     {
       "epoch": 2.416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.4169,
+      "grad_norm": 2.2128732204437256,
+      "learning_rate": 0.00044879999999999996,
+      "loss": 2.4414,
       "step": 1510
     },
     {
       "epoch": 2.4176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 0.8890330195426941,
+      "learning_rate": 0.0004491,
+      "loss": 2.0152,
       "step": 1511
     },
     {
       "epoch": 2.4192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.7872178554534912,
+      "learning_rate": 0.0004494,
+      "loss": 2.2456,
       "step": 1512
     },
     {
       "epoch": 2.4208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 5.388312816619873,
+      "learning_rate": 0.0004497,
+      "loss": 2.2783,
       "step": 1513
     },
     {
       "epoch": 2.4224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.207627534866333,
+      "learning_rate": 0.00045,
+      "loss": 2.2925,
       "step": 1514
     },
     {
       "epoch": 2.424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.7061299085617065,
+      "learning_rate": 0.00045029999999999994,
+      "loss": 2.0897,
       "step": 1515
     },
     {
       "epoch": 2.4256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.343909978866577,
+      "learning_rate": 0.00045059999999999995,
+      "loss": 2.2123,
       "step": 1516
     },
     {
       "epoch": 2.4272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.4407389163970947,
+      "learning_rate": 0.0004508999999999999,
+      "loss": 2.0337,
       "step": 1517
     },
     {
       "epoch": 2.4288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.6448605060577393,
+      "learning_rate": 0.00045119999999999996,
+      "loss": 2.0988,
       "step": 1518
     },
     {
       "epoch": 2.4304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.45,
+      "grad_norm": 2.379830837249756,
+      "learning_rate": 0.0004514999999999999,
+      "loss": 1.9966,
       "step": 1519
     },
     {
       "epoch": 2.432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.3262,
+      "grad_norm": 0.9263246059417725,
+      "learning_rate": 0.0004518,
+      "loss": 2.1314,
       "step": 1520
     },
     {
       "epoch": 2.4336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.748016595840454,
+      "learning_rate": 0.00045209999999999993,
+      "loss": 2.1768,
       "step": 1521
     },
     {
       "epoch": 2.4352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.4713168144226074,
+      "learning_rate": 0.00045239999999999994,
+      "loss": 2.1775,
       "step": 1522
     },
     {
       "epoch": 2.4368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.4357311725616455,
+      "learning_rate": 0.00045269999999999994,
+      "loss": 2.2528,
       "step": 1523
     },
     {
       "epoch": 2.4384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.8041242361068726,
+      "learning_rate": 0.00045299999999999995,
+      "loss": 2.082,
       "step": 1524
     },
     {
       "epoch": 2.44,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.257370948791504,
+      "learning_rate": 0.0004532999999999999,
+      "loss": 2.0602,
       "step": 1525
     },
     {
       "epoch": 2.4416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 0.9275694489479065,
+      "learning_rate": 0.00045359999999999997,
+      "loss": 2.0142,
       "step": 1526
     },
     {
       "epoch": 2.4432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 0.8489426970481873,
+      "learning_rate": 0.0004538999999999999,
+      "loss": 2.1113,
       "step": 1527
     },
     {
       "epoch": 2.4448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.2062907218933105,
+      "learning_rate": 0.0004542,
+      "loss": 2.1288,
       "step": 1528
     },
     {
       "epoch": 2.4464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 2.1949,
+      "grad_norm": 0.4519451856613159,
+      "learning_rate": 0.00045449999999999993,
+      "loss": 2.1187,
       "step": 1529
     },
     {
       "epoch": 2.448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 0.757920503616333,
+      "learning_rate": 0.00045479999999999994,
+      "loss": 2.1579,
       "step": 1530
     },
     {
       "epoch": 2.4496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.3224443197250366,
+      "learning_rate": 0.00045509999999999995,
+      "loss": 1.9504,
       "step": 1531
     },
     {
       "epoch": 2.4512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.6802444458007812,
+      "learning_rate": 0.00045539999999999996,
+      "loss": 2.1485,
       "step": 1532
     },
     {
       "epoch": 2.4528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.8263375759124756,
+      "learning_rate": 0.0004556999999999999,
+      "loss": 2.3139,
       "step": 1533
     },
     {
       "epoch": 2.4544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.8222333192825317,
+      "learning_rate": 0.00045599999999999997,
+      "loss": 2.1328,
       "step": 1534
     },
     {
       "epoch": 2.456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 1.3476057052612305,
+      "learning_rate": 0.0004562999999999999,
+      "loss": 2.0043,
       "step": 1535
     },
     {
       "epoch": 2.4576000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.153061866760254,
+      "learning_rate": 0.0004566,
+      "loss": 2.2033,
       "step": 1536
     },
     {
       "epoch": 2.4592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 1.2771,
+      "grad_norm": 1.0744353532791138,
+      "learning_rate": 0.00045689999999999994,
+      "loss": 2.115,
       "step": 1537
     },
     {
       "epoch": 2.4608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.4500582218170166,
+      "learning_rate": 0.00045719999999999995,
+      "loss": 2.4472,
       "step": 1538
     },
     {
       "epoch": 2.4624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 2.2745625972747803,
+      "learning_rate": 0.00045749999999999995,
+      "loss": 2.2543,
       "step": 1539
     },
     {
       "epoch": 2.464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005744,
-      "loss": 0.0,
+      "grad_norm": 3.1885974407196045,
+      "learning_rate": 0.00045779999999999996,
+      "loss": 2.269,
       "step": 1540
     },
     {
       "epoch": 2.4656000000000002,
-      "grad_norm": 0.0,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.7204,
+      "grad_norm": 3.2963569164276123,
+      "learning_rate": 0.0004580999999999999,
+      "loss": 2.2571,
       "step": 1541
     },
     {
       "epoch": 2.4672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.700000762939453,
+      "learning_rate": 0.0004584,
+      "loss": 2.4741,
       "step": 1542
     },
     {
       "epoch": 2.4688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.1224586963653564,
+      "learning_rate": 0.00045869999999999993,
+      "loss": 2.5694,
       "step": 1543
     },
     {
       "epoch": 2.4704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 1.2716,
+      "grad_norm": 1.3228379487991333,
+      "learning_rate": 0.000459,
+      "loss": 2.4327,
       "step": 1544
     },
     {
       "epoch": 2.472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.0606297254562378,
+      "learning_rate": 0.00045929999999999994,
+      "loss": 2.5588,
       "step": 1545
     },
     {
       "epoch": 2.4736000000000002,
       "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "learning_rate": 0.00045929999999999994,
+      "loss": 2.2503,
       "step": 1546
     },
     {
       "epoch": 2.4752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.2446088790893555,
+      "learning_rate": 0.00045959999999999995,
+      "loss": 2.4258,
       "step": 1547
     },
     {
       "epoch": 2.4768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.7627031803131104,
+      "learning_rate": 0.00045989999999999996,
+      "loss": 3.0056,
       "step": 1548
     },
     {
       "epoch": 2.4784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.4006,
+      "grad_norm": 1.893491506576538,
+      "learning_rate": 0.00046019999999999996,
+      "loss": 2.4505,
       "step": 1549
     },
     {
       "epoch": 2.48,
       "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "learning_rate": 0.00046019999999999996,
+      "loss": 2.2567,
       "step": 1550
     },
     {
       "epoch": 2.4816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 8.74563217163086,
+      "learning_rate": 0.0004604999999999999,
+      "loss": 2.9547,
       "step": 1551
     },
     {
       "epoch": 2.4832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.7902604341506958,
+      "learning_rate": 0.0004608,
+      "loss": 2.4997,
       "step": 1552
     },
     {
       "epoch": 2.4848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.6337521076202393,
+      "learning_rate": 0.00046109999999999993,
+      "loss": 2.6721,
       "step": 1553
     },
     {
       "epoch": 2.4864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.6069,
+      "grad_norm": 2.686281681060791,
+      "learning_rate": 0.0004614,
+      "loss": 2.4642,
       "step": 1554
     },
     {
       "epoch": 2.488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.8505804538726807,
+      "learning_rate": 0.00046169999999999995,
+      "loss": 2.6323,
       "step": 1555
     },
     {
       "epoch": 2.4896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 4.638845443725586,
+      "learning_rate": 0.00046199999999999995,
+      "loss": 2.4094,
       "step": 1556
     },
     {
       "epoch": 2.4912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 5.030619144439697,
+      "learning_rate": 0.00046229999999999996,
+      "loss": 2.4617,
       "step": 1557
     },
     {
       "epoch": 2.4928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.355173349380493,
+      "learning_rate": 0.00046259999999999997,
+      "loss": 2.7822,
       "step": 1558
     },
     {
       "epoch": 2.4944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 3.260460138320923,
+      "learning_rate": 0.0004628999999999999,
+      "loss": 2.393,
       "step": 1559
     },
     {
       "epoch": 2.496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.24070143699646,
+      "learning_rate": 0.0004632,
+      "loss": 2.3995,
       "step": 1560
     },
     {
       "epoch": 2.4976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.825697422027588,
+      "learning_rate": 0.00046349999999999994,
+      "loss": 2.1871,
       "step": 1561
     },
     {
       "epoch": 2.4992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.4646861553192139,
+      "learning_rate": 0.0004638,
+      "loss": 2.5,
       "step": 1562
     },
     {
       "epoch": 2.5008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.9098280668258667,
+      "learning_rate": 0.00046409999999999995,
+      "loss": 2.3278,
       "step": 1563
     },
     {
       "epoch": 2.5023999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.2409300804138184,
+      "learning_rate": 0.00046439999999999996,
+      "loss": 2.2171,
       "step": 1564
     },
     {
       "epoch": 2.504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.769909381866455,
+      "learning_rate": 0.00046469999999999997,
+      "loss": 2.1868,
       "step": 1565
     },
     {
       "epoch": 2.5056000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7229076623916626,
+      "learning_rate": 0.00046499999999999997,
+      "loss": 2.4986,
       "step": 1566
     },
     {
       "epoch": 2.5072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 3.2638115882873535,
+      "learning_rate": 0.0004652999999999999,
+      "loss": 2.4535,
       "step": 1567
     },
     {
       "epoch": 2.5088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.0078232288360596,
+      "learning_rate": 0.0004656,
+      "loss": 2.3516,
       "step": 1568
     },
     {
       "epoch": 2.5103999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.9068163633346558,
+      "learning_rate": 0.00046589999999999994,
+      "loss": 2.2269,
       "step": 1569
     },
     {
       "epoch": 2.512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7834655046463013,
+      "learning_rate": 0.00046619999999999995,
+      "loss": 2.1135,
       "step": 1570
     },
     {
       "epoch": 2.5136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.9387535452842712,
+      "learning_rate": 0.00046649999999999996,
+      "loss": 2.3821,
       "step": 1571
     },
     {
       "epoch": 2.5152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.49048078060150146,
+      "learning_rate": 0.00046679999999999996,
+      "loss": 2.2074,
       "step": 1572
     },
     {
       "epoch": 2.5168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7723761200904846,
+      "learning_rate": 0.0004670999999999999,
+      "loss": 2.2854,
       "step": 1573
     },
     {
       "epoch": 2.5183999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7409223318099976,
+      "learning_rate": 0.0004674,
+      "loss": 2.3194,
       "step": 1574
     },
     {
       "epoch": 2.52,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.611025333404541,
+      "learning_rate": 0.00046769999999999993,
+      "loss": 2.1594,
       "step": 1575
     },
     {
       "epoch": 2.5216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.4206929206848145,
+      "learning_rate": 0.000468,
+      "loss": 2.2845,
       "step": 1576
     },
     {
       "epoch": 2.5232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.7586792707443237,
+      "learning_rate": 0.00046829999999999994,
+      "loss": 2.241,
       "step": 1577
     },
     {
       "epoch": 2.5248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.6851300001144409,
+      "learning_rate": 0.00046859999999999995,
+      "loss": 2.1471,
       "step": 1578
     },
     {
       "epoch": 2.5263999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.724346935749054,
+      "learning_rate": 0.00046889999999999996,
+      "loss": 2.3002,
       "step": 1579
     },
     {
       "epoch": 2.528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.9342684745788574,
+      "learning_rate": 0.00046919999999999997,
+      "loss": 2.3658,
       "step": 1580
     },
     {
       "epoch": 2.5296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.6981943249702454,
+      "learning_rate": 0.0004694999999999999,
+      "loss": 2.197,
       "step": 1581
     },
     {
       "epoch": 2.5312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.4863,
+      "grad_norm": 1.980246901512146,
+      "learning_rate": 0.0004698,
+      "loss": 2.1856,
       "step": 1582
     },
     {
       "epoch": 2.5328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.1717616319656372,
+      "learning_rate": 0.00047009999999999993,
+      "loss": 2.1176,
       "step": 1583
     },
     {
       "epoch": 2.5343999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.5251626968383789,
+      "learning_rate": 0.0004704,
+      "loss": 2.2425,
       "step": 1584
     },
     {
       "epoch": 2.536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.7471,
+      "grad_norm": 0.5263369679450989,
+      "learning_rate": 0.00047069999999999995,
+      "loss": 2.4217,
       "step": 1585
     },
     {
       "epoch": 2.5376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.442,
+      "grad_norm": 1.7123208045959473,
+      "learning_rate": 0.00047099999999999996,
+      "loss": 2.3168,
       "step": 1586
     },
     {
       "epoch": 2.5392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.5260848999023438,
+      "learning_rate": 0.00047129999999999996,
+      "loss": 2.3844,
       "step": 1587
     },
     {
       "epoch": 2.5408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.0098893642425537,
+      "learning_rate": 0.00047159999999999997,
+      "loss": 2.2402,
       "step": 1588
     },
     {
       "epoch": 2.5423999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.457606554031372,
+      "learning_rate": 0.0004718999999999999,
+      "loss": 2.1904,
       "step": 1589
     },
     {
       "epoch": 2.544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.0802197456359863,
+      "learning_rate": 0.0004722,
+      "loss": 2.2182,
       "step": 1590
     },
     {
       "epoch": 2.5456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.5586844682693481,
+      "learning_rate": 0.00047249999999999994,
+      "loss": 2.2742,
       "step": 1591
     },
     {
       "epoch": 2.5472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.6091138124465942,
+      "learning_rate": 0.0004728,
+      "loss": 2.1756,
       "step": 1592
     },
     {
       "epoch": 2.5488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.9087393283843994,
+      "learning_rate": 0.00047309999999999995,
+      "loss": 2.1531,
       "step": 1593
     },
     {
       "epoch": 2.5504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 1.294,
+      "grad_norm": 1.046461582183838,
+      "learning_rate": 0.00047339999999999996,
+      "loss": 2.3781,
       "step": 1594
     },
     {
       "epoch": 2.552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.0787,
+      "grad_norm": 1.6911176443099976,
+      "learning_rate": 0.00047369999999999997,
+      "loss": 2.2942,
       "step": 1595
     },
     {
       "epoch": 2.5536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 1.1777,
+      "grad_norm": 2.0928030014038086,
+      "learning_rate": 0.000474,
+      "loss": 2.6092,
       "step": 1596
     },
     {
       "epoch": 2.5552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 1.2794,
+      "grad_norm": 3.662987232208252,
+      "learning_rate": 0.00047429999999999993,
+      "loss": 2.9461,
       "step": 1597
     },
     {
       "epoch": 2.5568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.8414223194122314,
+      "learning_rate": 0.0004746,
+      "loss": 2.5515,
       "step": 1598
     },
     {
       "epoch": 2.5584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 2.5247,
+      "grad_norm": 2.544562816619873,
+      "learning_rate": 0.00047489999999999994,
+      "loss": 2.6383,
       "step": 1599
     },
     {
       "epoch": 2.56,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
-      "step": 1600
-    },
-    {
-      "epoch": 2.56,
-      "eval_cer": 0.6112169649619325,
-      "eval_loss": 2.448988676071167,
-      "eval_runtime": 158.8119,
-      "eval_samples_per_second": 19.747,
-      "eval_steps_per_second": 1.234,
-      "eval_wer": 0.8888198114185059,
+      "grad_norm": 1.1964010000228882,
+      "learning_rate": 0.0004752,
+      "loss": 2.6473,
       "step": 1600
     },
     {
       "epoch": 2.5616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.2365444898605347,
+      "learning_rate": 0.00047549999999999996,
+      "loss": 2.6351,
       "step": 1601
     },
     {
       "epoch": 2.5632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.1067428588867188,
+      "learning_rate": 0.00047579999999999996,
+      "loss": 2.521,
       "step": 1602
     },
     {
       "epoch": 2.5648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 4.032146453857422,
+      "learning_rate": 0.00047609999999999997,
+      "loss": 2.7045,
       "step": 1603
     },
     {
       "epoch": 2.5664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.8868406414985657,
+      "learning_rate": 0.0004764,
+      "loss": 2.5629,
       "step": 1604
     },
     {
       "epoch": 2.568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.6565248966217041,
+      "learning_rate": 0.00047669999999999993,
+      "loss": 2.4577,
       "step": 1605
     },
     {
       "epoch": 2.5696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.5226905345916748,
+      "learning_rate": 0.000477,
+      "loss": 2.3847,
       "step": 1606
     },
     {
       "epoch": 2.5712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.1585739850997925,
+      "learning_rate": 0.00047729999999999995,
+      "loss": 2.5167,
       "step": 1607
     },
     {
       "epoch": 2.5728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 6.165229320526123,
+      "learning_rate": 0.0004776,
+      "loss": 2.4221,
       "step": 1608
     },
     {
       "epoch": 2.5744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 5.301511764526367,
+      "learning_rate": 0.00047789999999999996,
+      "loss": 2.2772,
       "step": 1609
     },
     {
       "epoch": 2.576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 5.172726631164551,
+      "learning_rate": 0.00047819999999999997,
+      "loss": 2.234,
       "step": 1610
     },
     {
       "epoch": 2.5776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 5.577144145965576,
+      "learning_rate": 0.0004785,
+      "loss": 2.3354,
       "step": 1611
     },
     {
       "epoch": 2.5792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 3.4328131675720215,
+      "learning_rate": 0.0004788,
+      "loss": 2.3838,
       "step": 1612
     },
     {
       "epoch": 2.5808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 3.080483913421631,
+      "learning_rate": 0.00047909999999999994,
+      "loss": 2.3324,
       "step": 1613
     },
     {
       "epoch": 2.5824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.9550327062606812,
+      "learning_rate": 0.0004794,
+      "loss": 2.0317,
       "step": 1614
     },
     {
       "epoch": 2.584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 1.1777,
+      "grad_norm": 1.0461270809173584,
+      "learning_rate": 0.00047969999999999995,
+      "loss": 2.0947,
       "step": 1615
     },
     {
       "epoch": 2.5856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.072354793548584,
+      "learning_rate": 0.00047999999999999996,
+      "loss": 2.0531,
       "step": 1616
     },
     {
       "epoch": 2.5872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.404462456703186,
+      "learning_rate": 0.00048029999999999997,
+      "loss": 2.0438,
       "step": 1617
     },
     {
       "epoch": 2.5888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.9748609066009521,
+      "learning_rate": 0.00048059999999999997,
+      "loss": 2.2188,
       "step": 1618
     },
     {
       "epoch": 2.5904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.9678696393966675,
+      "learning_rate": 0.0004808999999999999,
+      "loss": 2.0989,
       "step": 1619
     },
     {
       "epoch": 2.592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.535593032836914,
+      "learning_rate": 0.0004812,
+      "loss": 2.2185,
       "step": 1620
     },
     {
       "epoch": 2.5936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.916507363319397,
+      "learning_rate": 0.00048149999999999994,
+      "loss": 2.0964,
       "step": 1621
     },
     {
       "epoch": 2.5952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.4805487394332886,
+      "learning_rate": 0.0004818,
+      "loss": 2.1418,
       "step": 1622
     },
     {
       "epoch": 2.5968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.451712727546692,
+      "learning_rate": 0.00048209999999999995,
+      "loss": 1.9859,
       "step": 1623
     },
     {
       "epoch": 2.5984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.8536416292190552,
+      "learning_rate": 0.00048239999999999996,
+      "loss": 1.9668,
       "step": 1624
     },
     {
       "epoch": 2.6,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7038165926933289,
+      "learning_rate": 0.00048269999999999997,
+      "loss": 2.1383,
       "step": 1625
     },
     {
       "epoch": 2.6016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 1.0919,
+      "grad_norm": 2.2116916179656982,
+      "learning_rate": 0.000483,
+      "loss": 2.05,
       "step": 1626
     },
     {
       "epoch": 2.6032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7246533632278442,
+      "learning_rate": 0.00048329999999999993,
+      "loss": 2.0097,
       "step": 1627
     },
     {
       "epoch": 2.6048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.4604125022888184,
+      "learning_rate": 0.0004836,
+      "loss": 2.3906,
       "step": 1628
     },
     {
       "epoch": 2.6064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.2371268272399902,
+      "learning_rate": 0.00048389999999999994,
+      "loss": 2.003,
       "step": 1629
     },
     {
       "epoch": 2.608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.6980746984481812,
+      "learning_rate": 0.0004842,
+      "loss": 1.8967,
       "step": 1630
     },
     {
       "epoch": 2.6096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.7364569902420044,
+      "learning_rate": 0.00048449999999999996,
+      "loss": 1.9932,
       "step": 1631
     },
     {
       "epoch": 2.6112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 0.5320027470588684,
+      "learning_rate": 0.00048479999999999997,
+      "loss": 2.0484,
       "step": 1632
     },
     {
       "epoch": 2.6128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.290400981903076,
+      "learning_rate": 0.0004851,
+      "loss": 2.3368,
       "step": 1633
     },
     {
       "epoch": 2.6144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.922303318977356,
+      "learning_rate": 0.0004854,
+      "loss": 2.059,
       "step": 1634
     },
     {
       "epoch": 2.616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.3151150941848755,
+      "learning_rate": 0.00048569999999999993,
+      "loss": 2.2918,
       "step": 1635
     },
     {
       "epoch": 2.6176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.5266954898834229,
+      "learning_rate": 0.000486,
+      "loss": 2.1886,
       "step": 1636
     },
     {
       "epoch": 2.6192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.0405694246292114,
+      "learning_rate": 0.00048629999999999995,
+      "loss": 2.0374,
       "step": 1637
     },
     {
       "epoch": 2.6208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 2.313121795654297,
+      "learning_rate": 0.0004866,
+      "loss": 2.2178,
       "step": 1638
     },
     {
       "epoch": 2.6224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.2103170156478882,
+      "learning_rate": 0.00048689999999999996,
+      "loss": 2.0749,
       "step": 1639
     },
     {
       "epoch": 2.624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.970706820487976,
+      "learning_rate": 0.00048719999999999997,
+      "loss": 2.0443,
       "step": 1640
     },
     {
       "epoch": 2.6256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.3431305885314941,
+      "learning_rate": 0.0004875,
+      "loss": 1.8063,
       "step": 1641
     },
     {
       "epoch": 2.6272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.1013216972351074,
+      "learning_rate": 0.00048779999999999993,
+      "loss": 2.0287,
       "step": 1642
     },
     {
       "epoch": 2.6288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.6037465333938599,
+      "learning_rate": 0.00048809999999999994,
+      "loss": 1.9267,
       "step": 1643
     },
     {
       "epoch": 2.6304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005747999999999999,
-      "loss": 0.0,
+      "grad_norm": 1.6325836181640625,
+      "learning_rate": 0.0004883999999999999,
+      "loss": 1.9847,
       "step": 1644
     },
     {
       "epoch": 2.632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 2.4299,
+      "grad_norm": 0.6441347599029541,
+      "learning_rate": 0.0004887,
+      "loss": 1.8627,
       "step": 1645
     },
     {
       "epoch": 2.6336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.052411675453186,
+      "learning_rate": 0.000489,
+      "loss": 2.2007,
       "step": 1646
     },
     {
       "epoch": 2.6352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7423901557922363,
+      "learning_rate": 0.0004892999999999999,
+      "loss": 1.9574,
       "step": 1647
     },
     {
       "epoch": 2.6368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6799460053443909,
+      "learning_rate": 0.0004896,
+      "loss": 2.0761,
       "step": 1648
     },
     {
       "epoch": 2.6384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.829913854598999,
+      "learning_rate": 0.0004898999999999999,
+      "loss": 2.0512,
       "step": 1649
     },
     {
       "epoch": 2.64,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.362149715423584,
+      "learning_rate": 0.0004901999999999999,
+      "loss": 2.6416,
       "step": 1650
     },
     {
       "epoch": 2.6416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 31.19998550415039,
+      "learning_rate": 0.0004904999999999999,
+      "loss": 3.999,
       "step": 1651
     },
     {
       "epoch": 2.6432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.805262565612793,
+      "learning_rate": 0.0004907999999999999,
+      "loss": 2.3079,
       "step": 1652
     },
     {
       "epoch": 2.6448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1498773097991943,
+      "learning_rate": 0.0004911,
+      "loss": 2.5233,
       "step": 1653
     },
     {
       "epoch": 2.6464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1593884229660034,
+      "learning_rate": 0.0004913999999999999,
+      "loss": 2.505,
       "step": 1654
     },
     {
       "epoch": 2.648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.466228723526001,
+      "learning_rate": 0.0004917,
+      "loss": 2.3373,
       "step": 1655
     },
     {
       "epoch": 2.6496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7555994987487793,
+      "learning_rate": 0.0004919999999999999,
+      "loss": 2.07,
       "step": 1656
     },
     {
       "epoch": 2.6512000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5982134342193604,
+      "learning_rate": 0.0004923,
+      "loss": 2.0622,
       "step": 1657
     },
     {
       "epoch": 2.6528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.938797473907471,
+      "learning_rate": 0.0004925999999999999,
+      "loss": 2.2787,
       "step": 1658
     },
     {
       "epoch": 2.6544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7524287700653076,
+      "learning_rate": 0.0004929,
+      "loss": 2.1268,
       "step": 1659
     },
     {
       "epoch": 2.656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6103334426879883,
+      "learning_rate": 0.0004932,
+      "loss": 1.9649,
       "step": 1660
     },
     {
       "epoch": 2.6576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3654429912567139,
+      "learning_rate": 0.0004935,
+      "loss": 2.1234,
       "step": 1661
     },
     {
       "epoch": 2.6592000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2061095237731934,
+      "learning_rate": 0.0004938,
+      "loss": 2.073,
       "step": 1662
     },
     {
       "epoch": 2.6608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7076228857040405,
+      "learning_rate": 0.0004940999999999999,
+      "loss": 1.8841,
       "step": 1663
     },
     {
       "epoch": 2.6624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.660804033279419,
+      "learning_rate": 0.0004944,
+      "loss": 1.971,
       "step": 1664
     },
     {
       "epoch": 2.664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5702351331710815,
+      "learning_rate": 0.0004946999999999999,
+      "loss": 1.9213,
       "step": 1665
     },
     {
       "epoch": 2.6656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6872142553329468,
+      "learning_rate": 0.0004949999999999999,
+      "loss": 1.9347,
       "step": 1666
     },
     {
       "epoch": 2.6672000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4251344203948975,
+      "learning_rate": 0.0004953,
+      "loss": 2.3131,
       "step": 1667
     },
     {
       "epoch": 2.6688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.768301784992218,
+      "learning_rate": 0.0004955999999999999,
+      "loss": 2.0859,
       "step": 1668
     },
     {
       "epoch": 2.6704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.746718168258667,
+      "learning_rate": 0.0004959,
+      "loss": 2.052,
       "step": 1669
     },
     {
       "epoch": 2.672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.39943528175354,
+      "learning_rate": 0.0004961999999999999,
+      "loss": 1.8502,
       "step": 1670
     },
     {
       "epoch": 2.6736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1550344228744507,
+      "learning_rate": 0.0004965,
+      "loss": 2.1201,
       "step": 1671
     },
     {
       "epoch": 2.6752000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6754303574562073,
+      "learning_rate": 0.0004967999999999999,
+      "loss": 2.0508,
       "step": 1672
     },
     {
       "epoch": 2.6768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0880494117736816,
+      "learning_rate": 0.0004971,
+      "loss": 1.9002,
       "step": 1673
     },
     {
       "epoch": 2.6784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8594996333122253,
+      "learning_rate": 0.0004974,
+      "loss": 2.1085,
       "step": 1674
     },
     {
       "epoch": 2.68,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8045554757118225,
+      "learning_rate": 0.0004977,
+      "loss": 2.0949,
       "step": 1675
     },
     {
       "epoch": 2.6816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7858266830444336,
+      "learning_rate": 0.000498,
+      "loss": 1.7859,
       "step": 1676
     },
     {
       "epoch": 2.6832000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4103102684020996,
+      "learning_rate": 0.0004982999999999999,
+      "loss": 1.8545,
       "step": 1677
     },
     {
       "epoch": 2.6848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5115968585014343,
+      "learning_rate": 0.0004986,
+      "loss": 2.1158,
       "step": 1678
     },
     {
       "epoch": 2.6864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2175918817520142,
+      "learning_rate": 0.0004988999999999999,
+      "loss": 1.9361,
       "step": 1679
     },
     {
       "epoch": 2.6879999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6044435501098633,
+      "learning_rate": 0.0004991999999999999,
+      "loss": 1.8655,
       "step": 1680
     },
     {
       "epoch": 2.6896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8397135138511658,
+      "learning_rate": 0.0004994999999999999,
+      "loss": 1.9778,
       "step": 1681
     },
     {
       "epoch": 2.6912000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6573118567466736,
+      "learning_rate": 0.0004997999999999999,
+      "loss": 1.9161,
       "step": 1682
     },
     {
       "epoch": 2.6928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.799583911895752,
+      "learning_rate": 0.0005001,
+      "loss": 2.1068,
       "step": 1683
     },
     {
       "epoch": 2.6944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9398294687271118,
+      "learning_rate": 0.0005003999999999999,
+      "loss": 2.1357,
       "step": 1684
     },
     {
       "epoch": 2.6959999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.780285656452179,
+      "learning_rate": 0.0005007,
+      "loss": 2.1124,
       "step": 1685
     },
     {
       "epoch": 2.6976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7242156863212585,
+      "learning_rate": 0.0005009999999999999,
+      "loss": 2.1581,
       "step": 1686
     },
     {
       "epoch": 2.6992000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8511390089988708,
+      "learning_rate": 0.0005013,
+      "loss": 1.9856,
       "step": 1687
     },
     {
       "epoch": 2.7008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6015489101409912,
+      "learning_rate": 0.0005015999999999999,
+      "loss": 2.1612,
       "step": 1688
     },
     {
       "epoch": 2.7024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.49614596366882324,
+      "learning_rate": 0.0005019,
+      "loss": 1.9588,
       "step": 1689
     },
     {
       "epoch": 2.7039999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5268350839614868,
+      "learning_rate": 0.0005022,
+      "loss": 1.739,
       "step": 1690
     },
     {
       "epoch": 2.7056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3183175325393677,
+      "learning_rate": 0.0005025,
+      "loss": 2.3823,
       "step": 1691
     },
     {
       "epoch": 2.7072000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8698546290397644,
+      "learning_rate": 0.0005028,
+      "loss": 1.8146,
       "step": 1692
     },
     {
       "epoch": 2.7088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7403873205184937,
+      "learning_rate": 0.0005030999999999999,
+      "loss": 2.0503,
       "step": 1693
     },
     {
       "epoch": 2.7104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6730201840400696,
+      "learning_rate": 0.0005034,
+      "loss": 2.1787,
       "step": 1694
     },
     {
       "epoch": 2.7119999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9731574654579163,
+      "learning_rate": 0.0005036999999999999,
+      "loss": 1.7114,
       "step": 1695
     },
     {
       "epoch": 2.7136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7726271748542786,
+      "learning_rate": 0.0005039999999999999,
+      "loss": 1.8783,
       "step": 1696
     },
     {
       "epoch": 2.7152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9202337265014648,
+      "learning_rate": 0.0005043,
+      "loss": 2.3164,
       "step": 1697
     },
     {
       "epoch": 2.7168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9275617599487305,
+      "learning_rate": 0.0005045999999999999,
+      "loss": 2.359,
       "step": 1698
     },
     {
       "epoch": 2.7184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6160383224487305,
+      "learning_rate": 0.0005049,
+      "loss": 2.0426,
       "step": 1699
     },
     {
       "epoch": 2.7199999999999998,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 1700
-    },
-    {
-      "epoch": 2.7199999999999998,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 162.8269,
-      "eval_samples_per_second": 19.26,
-      "eval_steps_per_second": 1.204,
-      "eval_wer": 1.0,
+      "learning_rate": 0.0005049,
+      "loss": 2.2236,
       "step": 1700
     },
     {
       "epoch": 2.7216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.13432502746582,
+      "learning_rate": 0.0005051999999999999,
+      "loss": 3.2363,
       "step": 1701
     },
     {
       "epoch": 2.7232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4809963703155518,
+      "learning_rate": 0.0005055,
+      "loss": 2.4094,
       "step": 1702
     },
     {
       "epoch": 2.7248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4243474006652832,
+      "learning_rate": 0.0005057999999999999,
+      "loss": 2.2277,
       "step": 1703
     },
     {
       "epoch": 2.7264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8862740993499756,
+      "learning_rate": 0.0005061,
+      "loss": 2.4024,
       "step": 1704
     },
     {
       "epoch": 2.7279999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.541082382202148,
+      "learning_rate": 0.0005064,
+      "loss": 2.1635,
       "step": 1705
     },
     {
       "epoch": 2.7296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9146323204040527,
+      "learning_rate": 0.0005067,
+      "loss": 2.1451,
       "step": 1706
     },
     {
       "epoch": 2.7312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9531452655792236,
+      "learning_rate": 0.000507,
+      "loss": 2.1887,
       "step": 1707
     },
     {
       "epoch": 2.7328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.83717155456543,
+      "learning_rate": 0.0005073,
+      "loss": 2.6369,
       "step": 1708
     },
     {
       "epoch": 2.7344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.303669810295105,
+      "learning_rate": 0.0005076,
+      "loss": 2.0226,
       "step": 1709
     },
     {
       "epoch": 2.7359999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5467097163200378,
+      "learning_rate": 0.0005078999999999999,
+      "loss": 1.9454,
       "step": 1710
     },
     {
       "epoch": 2.7376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.554548740386963,
+      "learning_rate": 0.0005082,
+      "loss": 2.0932,
       "step": 1711
     },
     {
       "epoch": 2.7392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5141626596450806,
+      "learning_rate": 0.0005085,
+      "loss": 2.1044,
       "step": 1712
     },
     {
       "epoch": 2.7408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.388733148574829,
+      "learning_rate": 0.0005087999999999999,
+      "loss": 2.3155,
       "step": 1713
     },
     {
       "epoch": 2.7424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.090745210647583,
+      "learning_rate": 0.0005091,
+      "loss": 2.0168,
       "step": 1714
     },
     {
       "epoch": 2.7439999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5126121044158936,
+      "learning_rate": 0.0005093999999999999,
+      "loss": 1.988,
       "step": 1715
     },
     {
       "epoch": 2.7456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.547797918319702,
+      "learning_rate": 0.0005097,
+      "loss": 1.9305,
       "step": 1716
     },
     {
       "epoch": 2.7472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9627645611763,
+      "learning_rate": 0.0005099999999999999,
+      "loss": 1.9474,
       "step": 1717
     },
     {
       "epoch": 2.7488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.361758232116699,
+      "learning_rate": 0.0005103,
+      "loss": 2.0709,
       "step": 1718
     },
     {
       "epoch": 2.7504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1550334692001343,
+      "learning_rate": 0.0005105999999999999,
+      "loss": 1.8985,
       "step": 1719
     },
     {
       "epoch": 2.752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.120604157447815,
+      "learning_rate": 0.0005109,
+      "loss": 1.8786,
       "step": 1720
     },
     {
       "epoch": 2.7536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5634665489196777,
+      "learning_rate": 0.0005112,
+      "loss": 1.9776,
       "step": 1721
     },
     {
       "epoch": 2.7552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.949796676635742,
+      "learning_rate": 0.0005115,
+      "loss": 1.9238,
       "step": 1722
     },
     {
       "epoch": 2.7568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8489679098129272,
+      "learning_rate": 0.0005118,
+      "loss": 1.695,
       "step": 1723
     },
     {
       "epoch": 2.7584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8145275115966797,
+      "learning_rate": 0.0005120999999999999,
+      "loss": 2.0701,
       "step": 1724
     },
     {
       "epoch": 2.76,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7843981981277466,
+      "learning_rate": 0.0005124,
+      "loss": 1.7572,
       "step": 1725
     },
     {
       "epoch": 2.7616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4644101560115814,
+      "learning_rate": 0.0005126999999999999,
+      "loss": 1.7349,
       "step": 1726
     },
     {
       "epoch": 2.7632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9464445114135742,
+      "learning_rate": 0.0005129999999999999,
+      "loss": 1.9303,
       "step": 1727
     },
     {
       "epoch": 2.7648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.003891110420227,
+      "learning_rate": 0.0005133,
+      "loss": 1.7827,
       "step": 1728
     },
     {
       "epoch": 2.7664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8749507665634155,
+      "learning_rate": 0.0005135999999999999,
+      "loss": 1.941,
       "step": 1729
     },
     {
       "epoch": 2.768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8504045009613037,
+      "learning_rate": 0.0005139,
+      "loss": 2.2031,
       "step": 1730
     },
     {
       "epoch": 2.7696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7272230982780457,
+      "learning_rate": 0.0005141999999999999,
+      "loss": 2.2033,
       "step": 1731
     },
     {
       "epoch": 2.7712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5356547832489014,
+      "learning_rate": 0.0005145,
+      "loss": 1.7236,
       "step": 1732
     },
     {
       "epoch": 2.7728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2203516960144043,
+      "learning_rate": 0.0005147999999999999,
+      "loss": 2.0358,
       "step": 1733
     },
     {
       "epoch": 2.7744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8269984126091003,
+      "learning_rate": 0.0005151,
+      "loss": 2.1528,
       "step": 1734
     },
     {
       "epoch": 2.776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1500989198684692,
+      "learning_rate": 0.0005154,
+      "loss": 1.9726,
       "step": 1735
     },
     {
       "epoch": 2.7776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8158721923828125,
+      "learning_rate": 0.0005157,
+      "loss": 2.0221,
       "step": 1736
     },
     {
       "epoch": 2.7792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5364477634429932,
+      "learning_rate": 0.000516,
+      "loss": 1.9459,
       "step": 1737
     },
     {
       "epoch": 2.7808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5089154839515686,
+      "learning_rate": 0.0005163,
+      "loss": 1.842,
       "step": 1738
     },
     {
       "epoch": 2.7824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3278019428253174,
+      "learning_rate": 0.0005166,
+      "loss": 2.0213,
       "step": 1739
     },
     {
       "epoch": 2.784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6192806959152222,
+      "learning_rate": 0.0005168999999999999,
+      "loss": 1.7591,
       "step": 1740
     },
     {
       "epoch": 2.7856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6816684007644653,
+      "learning_rate": 0.0005172,
+      "loss": 1.9409,
       "step": 1741
     },
     {
       "epoch": 2.7872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6508601307868958,
+      "learning_rate": 0.0005175,
+      "loss": 1.8534,
       "step": 1742
     },
     {
       "epoch": 2.7888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8660383224487305,
+      "learning_rate": 0.0005177999999999999,
+      "loss": 1.9621,
       "step": 1743
     },
     {
       "epoch": 2.7904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5654451251029968,
+      "learning_rate": 0.0005181,
+      "loss": 1.8793,
       "step": 1744
     },
     {
       "epoch": 2.792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.010545015335083,
+      "learning_rate": 0.0005183999999999999,
+      "loss": 1.8519,
       "step": 1745
     },
     {
       "epoch": 2.7936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6233418583869934,
+      "learning_rate": 0.0005187,
+      "loss": 2.305,
       "step": 1746
     },
     {
       "epoch": 2.7952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8341806530952454,
+      "learning_rate": 0.0005189999999999999,
+      "loss": 2.2414,
       "step": 1747
     },
     {
       "epoch": 2.7968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7348121404647827,
+      "learning_rate": 0.0005193,
+      "loss": 1.8707,
       "step": 1748
     },
     {
       "epoch": 2.7984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8103607892990112,
+      "learning_rate": 0.0005195999999999999,
+      "loss": 2.0222,
       "step": 1749
     },
     {
       "epoch": 2.8,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6983300447463989,
+      "learning_rate": 0.0005199,
+      "loss": 2.1295,
       "step": 1750
     },
     {
       "epoch": 2.8016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.782421588897705,
+      "learning_rate": 0.0005202,
+      "loss": 2.5126,
       "step": 1751
     },
     {
       "epoch": 2.8032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.138143539428711,
+      "learning_rate": 0.0005205,
+      "loss": 2.1409,
       "step": 1752
     },
     {
       "epoch": 2.8048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9656838178634644,
+      "learning_rate": 0.0005208,
+      "loss": 2.3417,
       "step": 1753
     },
     {
       "epoch": 2.8064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.654718816280365,
+      "learning_rate": 0.0005211,
+      "loss": 2.114,
       "step": 1754
     },
     {
       "epoch": 2.808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.243199348449707,
+      "learning_rate": 0.0005214,
+      "loss": 2.5875,
       "step": 1755
     },
     {
       "epoch": 2.8096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9750616550445557,
+      "learning_rate": 0.0005216999999999999,
+      "loss": 2.0969,
       "step": 1756
     },
     {
       "epoch": 2.8112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2397475242614746,
+      "learning_rate": 0.000522,
+      "loss": 2.0965,
       "step": 1757
     },
     {
       "epoch": 2.8128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9542648792266846,
+      "learning_rate": 0.0005223,
+      "loss": 2.0542,
       "step": 1758
     },
     {
       "epoch": 2.8144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8388017416000366,
+      "learning_rate": 0.0005225999999999999,
+      "loss": 2.2752,
       "step": 1759
     },
     {
       "epoch": 2.816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9828288555145264,
+      "learning_rate": 0.0005229,
+      "loss": 1.901,
       "step": 1760
     },
     {
       "epoch": 2.8176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.787229061126709,
+      "learning_rate": 0.0005231999999999999,
+      "loss": 1.7862,
       "step": 1761
     },
     {
       "epoch": 2.8192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.40390992164611816,
+      "learning_rate": 0.0005235,
+      "loss": 1.9124,
       "step": 1762
     },
     {
       "epoch": 2.8208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3556129932403564,
+      "learning_rate": 0.0005237999999999999,
+      "loss": 2.0211,
       "step": 1763
     },
     {
       "epoch": 2.8224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4864728450775146,
+      "learning_rate": 0.0005241,
+      "loss": 1.775,
       "step": 1764
     },
     {
       "epoch": 2.824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8218860626220703,
+      "learning_rate": 0.0005244,
+      "loss": 1.8258,
       "step": 1765
     },
     {
       "epoch": 2.8256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.097517967224121,
+      "learning_rate": 0.0005247,
+      "loss": 1.8297,
       "step": 1766
     },
     {
       "epoch": 2.8272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.49936580657959,
+      "learning_rate": 0.000525,
+      "loss": 2.0836,
       "step": 1767
     },
     {
       "epoch": 2.8288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7731903791427612,
+      "learning_rate": 0.0005252999999999999,
+      "loss": 1.7411,
       "step": 1768
     },
     {
       "epoch": 2.8304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8479838967323303,
+      "learning_rate": 0.0005256,
+      "loss": 1.6936,
       "step": 1769
     },
     {
       "epoch": 2.832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.757756233215332,
+      "learning_rate": 0.0005258999999999999,
+      "loss": 1.9003,
       "step": 1770
     },
     {
       "epoch": 2.8336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8400256633758545,
+      "learning_rate": 0.0005262,
+      "loss": 1.7133,
       "step": 1771
     },
     {
       "epoch": 2.8352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8254936933517456,
+      "learning_rate": 0.0005265,
+      "loss": 1.8496,
       "step": 1772
     },
     {
       "epoch": 2.8368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5556833148002625,
+      "learning_rate": 0.0005267999999999999,
+      "loss": 1.8549,
       "step": 1773
     },
     {
       "epoch": 2.8384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6342326402664185,
+      "learning_rate": 0.0005271,
+      "loss": 2.008,
       "step": 1774
     },
     {
       "epoch": 2.84,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6860460638999939,
+      "learning_rate": 0.0005273999999999999,
+      "loss": 1.7957,
       "step": 1775
     },
     {
       "epoch": 2.8416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0764575004577637,
+      "learning_rate": 0.0005276999999999999,
+      "loss": 2.0447,
       "step": 1776
     },
     {
       "epoch": 2.8432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4549970030784607,
+      "learning_rate": 0.0005279999999999999,
+      "loss": 2.0276,
       "step": 1777
     },
     {
       "epoch": 2.8448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5434354543685913,
+      "learning_rate": 0.0005282999999999999,
+      "loss": 1.9484,
       "step": 1778
     },
     {
       "epoch": 2.8464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8522821664810181,
+      "learning_rate": 0.0005286,
+      "loss": 1.7575,
       "step": 1779
     },
     {
       "epoch": 2.848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.750393271446228,
+      "learning_rate": 0.0005288999999999999,
+      "loss": 1.9477,
       "step": 1780
     },
     {
       "epoch": 2.8496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7382807731628418,
+      "learning_rate": 0.0005292,
+      "loss": 1.8868,
       "step": 1781
     },
     {
       "epoch": 2.8512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1927576065063477,
+      "learning_rate": 0.0005294999999999999,
+      "loss": 2.0142,
       "step": 1782
     },
     {
       "epoch": 2.8528000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0110162496566772,
+      "learning_rate": 0.0005298,
+      "loss": 2.0582,
       "step": 1783
     },
     {
       "epoch": 2.8544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6020126342773438,
+      "learning_rate": 0.0005300999999999999,
+      "loss": 1.7602,
       "step": 1784
     },
     {
       "epoch": 2.856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7490088939666748,
+      "learning_rate": 0.0005304,
+      "loss": 1.8641,
       "step": 1785
     },
     {
       "epoch": 2.8576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5195388793945312,
+      "learning_rate": 0.0005306999999999999,
+      "loss": 1.7468,
       "step": 1786
     },
     {
       "epoch": 2.8592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7462009191513062,
+      "learning_rate": 0.000531,
+      "loss": 1.9464,
       "step": 1787
     },
     {
       "epoch": 2.8608000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5070817470550537,
+      "learning_rate": 0.0005313,
+      "loss": 1.949,
       "step": 1788
     },
     {
       "epoch": 2.8624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6855885982513428,
+      "learning_rate": 0.0005315999999999999,
+      "loss": 2.1353,
       "step": 1789
     },
     {
       "epoch": 2.864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7518872618675232,
+      "learning_rate": 0.0005319,
+      "loss": 2.0692,
       "step": 1790
     },
     {
       "epoch": 2.8656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1828621625900269,
+      "learning_rate": 0.0005321999999999999,
+      "loss": 1.9593,
       "step": 1791
     },
     {
       "epoch": 2.8672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2454311847686768,
+      "learning_rate": 0.0005324999999999999,
+      "loss": 1.8795,
       "step": 1792
     },
     {
       "epoch": 2.8688000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3150829076766968,
+      "learning_rate": 0.0005327999999999999,
+      "loss": 2.0185,
       "step": 1793
     },
     {
       "epoch": 2.8704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9219515323638916,
+      "learning_rate": 0.0005330999999999999,
+      "loss": 2.183,
       "step": 1794
     },
     {
       "epoch": 2.872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1829967498779297,
+      "learning_rate": 0.0005334,
+      "loss": 2.3794,
       "step": 1795
     },
     {
       "epoch": 2.8736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.068989872932434,
+      "learning_rate": 0.0005336999999999999,
+      "loss": 1.9122,
       "step": 1796
     },
     {
       "epoch": 2.8752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4533793926239014,
+      "learning_rate": 0.000534,
+      "loss": 1.7805,
       "step": 1797
     },
     {
       "epoch": 2.8768000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4632128477096558,
+      "learning_rate": 0.0005342999999999999,
+      "loss": 2.3252,
       "step": 1798
     },
     {
       "epoch": 2.8784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.011090636253357,
+      "learning_rate": 0.0005346,
+      "loss": 2.1687,
       "step": 1799
     },
     {
       "epoch": 2.88,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 1800
-    },
-    {
-      "epoch": 2.88,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 158.7253,
-      "eval_samples_per_second": 19.757,
-      "eval_steps_per_second": 1.235,
-      "eval_wer": 1.0,
+      "grad_norm": 1.1097999811172485,
+      "learning_rate": 0.0005348999999999999,
+      "loss": 2.2309,
       "step": 1800
     },
     {
       "epoch": 2.8816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.508502006530762,
+      "learning_rate": 0.0005352,
+      "loss": 2.6601,
       "step": 1801
     },
     {
       "epoch": 2.8832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.70348834991455,
+      "learning_rate": 0.0005355,
+      "loss": 2.6031,
       "step": 1802
     },
     {
       "epoch": 2.8848000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.40385365486145,
+      "learning_rate": 0.0005358,
+      "loss": 2.0591,
       "step": 1803
     },
     {
       "epoch": 2.8864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.063434600830078,
+      "learning_rate": 0.0005361,
+      "loss": 2.304,
       "step": 1804
     },
     {
       "epoch": 2.888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9008119106292725,
+      "learning_rate": 0.0005363999999999999,
+      "loss": 2.3875,
       "step": 1805
     },
     {
       "epoch": 2.8895999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3916290998458862,
+      "learning_rate": 0.0005367,
+      "loss": 1.9926,
       "step": 1806
     },
     {
       "epoch": 2.8912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7277209758758545,
+      "learning_rate": 0.0005369999999999999,
+      "loss": 1.9297,
       "step": 1807
     },
     {
       "epoch": 2.8928000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7167696952819824,
+      "learning_rate": 0.0005372999999999999,
+      "loss": 1.8832,
       "step": 1808
     },
     {
       "epoch": 2.8944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2191977500915527,
+      "learning_rate": 0.0005376,
+      "loss": 2.1223,
       "step": 1809
     },
     {
       "epoch": 2.896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.684467315673828,
+      "learning_rate": 0.0005378999999999999,
+      "loss": 2.4019,
       "step": 1810
     },
     {
       "epoch": 2.8975999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.767059326171875,
+      "learning_rate": 0.0005382,
+      "loss": 1.9166,
       "step": 1811
     },
     {
       "epoch": 2.8992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2353407144546509,
+      "learning_rate": 0.0005384999999999999,
+      "loss": 2.0115,
       "step": 1812
     },
     {
       "epoch": 2.9008000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.391242742538452,
+      "learning_rate": 0.0005388,
+      "loss": 1.7517,
       "step": 1813
     },
     {
       "epoch": 2.9024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6257333755493164,
+      "learning_rate": 0.0005390999999999999,
+      "loss": 1.7542,
       "step": 1814
     },
     {
       "epoch": 2.904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6803297996520996,
+      "learning_rate": 0.0005394,
+      "loss": 1.7067,
       "step": 1815
     },
     {
       "epoch": 2.9055999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7414671182632446,
+      "learning_rate": 0.0005396999999999999,
+      "loss": 1.7507,
       "step": 1816
     },
     {
       "epoch": 2.9072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1055442094802856,
+      "learning_rate": 0.00054,
+      "loss": 1.9703,
       "step": 1817
     },
     {
       "epoch": 2.9088000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4091672897338867,
+      "learning_rate": 0.0005403,
+      "loss": 1.9673,
       "step": 1818
     },
     {
       "epoch": 2.9104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8116196393966675,
+      "learning_rate": 0.0005405999999999999,
+      "loss": 1.7938,
       "step": 1819
     },
     {
       "epoch": 2.912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4093546867370605,
+      "learning_rate": 0.0005409,
+      "loss": 1.9355,
       "step": 1820
     },
     {
       "epoch": 2.9135999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8408976197242737,
+      "learning_rate": 0.0005411999999999999,
+      "loss": 2.0643,
       "step": 1821
     },
     {
       "epoch": 2.9152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.2744598388671875,
+      "learning_rate": 0.0005414999999999999,
+      "loss": 2.2347,
       "step": 1822
     },
     {
       "epoch": 2.9168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5432764291763306,
+      "learning_rate": 0.0005417999999999999,
+      "loss": 1.9725,
       "step": 1823
     },
     {
       "epoch": 2.9184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5486776232719421,
+      "learning_rate": 0.0005420999999999999,
+      "loss": 1.5897,
       "step": 1824
     },
     {
       "epoch": 2.92,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.742293119430542,
+      "learning_rate": 0.0005424,
+      "loss": 2.0866,
       "step": 1825
     },
     {
       "epoch": 2.9215999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.380513310432434,
+      "learning_rate": 0.0005426999999999999,
+      "loss": 1.7298,
       "step": 1826
     },
     {
       "epoch": 2.9232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8607488870620728,
+      "learning_rate": 0.000543,
+      "loss": 1.8805,
       "step": 1827
     },
     {
       "epoch": 2.9248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.006989598274231,
+      "learning_rate": 0.0005432999999999999,
+      "loss": 2.0631,
       "step": 1828
     },
     {
       "epoch": 2.9264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6280760765075684,
+      "learning_rate": 0.0005436,
+      "loss": 1.8232,
       "step": 1829
     },
     {
       "epoch": 2.928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5118163228034973,
+      "learning_rate": 0.0005438999999999999,
+      "loss": 2.0281,
       "step": 1830
     },
     {
       "epoch": 2.9295999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9178122282028198,
+      "learning_rate": 0.0005442,
+      "loss": 1.7694,
       "step": 1831
     },
     {
       "epoch": 2.9312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.245267629623413,
+      "learning_rate": 0.0005445,
+      "loss": 2.2356,
       "step": 1832
     },
     {
       "epoch": 2.9328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5646647214889526,
+      "learning_rate": 0.0005448,
+      "loss": 2.111,
       "step": 1833
     },
     {
       "epoch": 2.9344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.696097493171692,
+      "learning_rate": 0.0005451,
+      "loss": 2.0895,
       "step": 1834
     },
     {
       "epoch": 2.936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9454824328422546,
+      "learning_rate": 0.0005453999999999999,
+      "loss": 1.9672,
       "step": 1835
     },
     {
       "epoch": 2.9375999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8050206899642944,
+      "learning_rate": 0.0005457,
+      "loss": 1.8222,
       "step": 1836
     },
     {
       "epoch": 2.9392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.688675880432129,
+      "learning_rate": 0.0005459999999999999,
+      "loss": 1.9278,
       "step": 1837
     },
     {
       "epoch": 2.9408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3390727043151855,
+      "learning_rate": 0.0005462999999999999,
+      "loss": 2.0807,
       "step": 1838
     },
     {
       "epoch": 2.9424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0786921977996826,
+      "learning_rate": 0.0005466,
+      "loss": 2.1307,
       "step": 1839
     },
     {
       "epoch": 2.944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0735629796981812,
+      "learning_rate": 0.0005468999999999999,
+      "loss": 1.7635,
       "step": 1840
     },
     {
       "epoch": 2.9455999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2507575750350952,
+      "learning_rate": 0.0005472,
+      "loss": 1.812,
       "step": 1841
     },
     {
       "epoch": 2.9472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5095469355583191,
+      "learning_rate": 0.0005474999999999999,
+      "loss": 1.9677,
       "step": 1842
     },
     {
       "epoch": 2.9488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0531517267227173,
+      "learning_rate": 0.0005478,
+      "loss": 1.8112,
       "step": 1843
     },
     {
       "epoch": 2.9504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.995157241821289,
+      "learning_rate": 0.0005480999999999999,
+      "loss": 2.3632,
       "step": 1844
     },
     {
       "epoch": 2.952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5749807953834534,
+      "learning_rate": 0.0005484,
+      "loss": 2.0844,
       "step": 1845
     },
     {
       "epoch": 2.9536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7241598963737488,
+      "learning_rate": 0.0005487,
+      "loss": 2.0575,
       "step": 1846
     },
     {
       "epoch": 2.9552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.396909475326538,
+      "learning_rate": 0.000549,
+      "loss": 2.2126,
       "step": 1847
     },
     {
       "epoch": 2.9568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5103518962860107,
+      "learning_rate": 0.0005493,
+      "loss": 2.276,
       "step": 1848
     },
     {
       "epoch": 2.9584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.392578601837158,
+      "learning_rate": 0.0005496,
+      "loss": 2.3884,
       "step": 1849
     },
     {
       "epoch": 2.96,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7753760814666748,
+      "learning_rate": 0.0005499,
+      "loss": 2.3444,
       "step": 1850
     },
     {
       "epoch": 2.9616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.469567060470581,
+      "learning_rate": 0.0005501999999999999,
+      "loss": 2.3639,
       "step": 1851
     },
     {
       "epoch": 2.9632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1051902770996094,
+      "learning_rate": 0.0005505,
+      "loss": 2.4397,
       "step": 1852
     },
     {
       "epoch": 2.9648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8053222894668579,
+      "learning_rate": 0.0005507999999999999,
+      "loss": 2.4935,
       "step": 1853
     },
     {
       "epoch": 2.9664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.687619924545288,
+      "learning_rate": 0.0005510999999999999,
+      "loss": 1.8974,
       "step": 1854
     },
     {
       "epoch": 2.968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.211838722229004,
+      "learning_rate": 0.0005514,
+      "loss": 1.8626,
       "step": 1855
     },
     {
       "epoch": 2.9696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.72635817527771,
+      "learning_rate": 0.0005516999999999999,
+      "loss": 1.9059,
       "step": 1856
     },
     {
       "epoch": 2.9712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.077843427658081,
+      "learning_rate": 0.000552,
+      "loss": 1.9353,
       "step": 1857
     },
     {
       "epoch": 2.9728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.494789123535156,
+      "learning_rate": 0.0005522999999999999,
+      "loss": 2.1245,
       "step": 1858
     },
     {
       "epoch": 2.9744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.749758243560791,
+      "learning_rate": 0.0005526,
+      "loss": 1.8147,
       "step": 1859
     },
     {
       "epoch": 2.976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0033164024353027,
+      "learning_rate": 0.0005528999999999999,
+      "loss": 1.8362,
       "step": 1860
     },
     {
       "epoch": 2.9776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6518697738647461,
+      "learning_rate": 0.0005532,
+      "loss": 2.2349,
       "step": 1861
     },
     {
       "epoch": 2.9792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.695722222328186,
+      "learning_rate": 0.0005535,
+      "loss": 1.7574,
       "step": 1862
     },
     {
       "epoch": 2.9808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5895578861236572,
+      "learning_rate": 0.0005538,
+      "loss": 2.0763,
       "step": 1863
     },
     {
       "epoch": 2.9824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0296926498413086,
+      "learning_rate": 0.0005541,
+      "loss": 1.9809,
       "step": 1864
     },
     {
       "epoch": 2.984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6186490654945374,
+      "learning_rate": 0.0005544,
+      "loss": 1.8883,
       "step": 1865
     },
     {
       "epoch": 2.9856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3189398050308228,
+      "learning_rate": 0.0005547,
+      "loss": 1.8607,
       "step": 1866
     },
     {
       "epoch": 2.9872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6981890201568604,
+      "learning_rate": 0.0005549999999999999,
+      "loss": 1.9978,
       "step": 1867
     },
     {
       "epoch": 2.9888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.823566734790802,
+      "learning_rate": 0.0005552999999999999,
+      "loss": 1.744,
       "step": 1868
     },
     {
       "epoch": 2.9904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6368497610092163,
+      "learning_rate": 0.0005556,
+      "loss": 2.0121,
       "step": 1869
     },
     {
       "epoch": 2.992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7579899430274963,
+      "learning_rate": 0.0005558999999999999,
+      "loss": 1.778,
       "step": 1870
     },
     {
       "epoch": 2.9936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4312794208526611,
+      "learning_rate": 0.0005562,
+      "loss": 2.0611,
       "step": 1871
     },
     {
       "epoch": 2.9952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6638422608375549,
+      "learning_rate": 0.0005564999999999999,
+      "loss": 1.9862,
       "step": 1872
     },
     {
       "epoch": 2.9968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5944936275482178,
+      "learning_rate": 0.0005568,
+      "loss": 2.1282,
       "step": 1873
     },
     {
       "epoch": 2.9984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.530720829963684,
+      "learning_rate": 0.0005570999999999999,
+      "loss": 1.972,
       "step": 1874
     },
     {
       "epoch": 3.0,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1473948955535889,
+      "learning_rate": 0.0005574,
+      "loss": 2.3039,
       "step": 1875
     },
     {
       "epoch": 3.0016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 15.021516799926758,
+      "learning_rate": 0.0005577,
+      "loss": 3.0943,
       "step": 1876
     },
     {
       "epoch": 3.0032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.774639129638672,
+      "learning_rate": 0.000558,
+      "loss": 2.3365,
       "step": 1877
     },
     {
       "epoch": 3.0048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.604449272155762,
+      "learning_rate": 0.0005583,
+      "loss": 2.5506,
       "step": 1878
     },
     {
       "epoch": 3.0064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6512893438339233,
+      "learning_rate": 0.0005586,
+      "loss": 2.2356,
       "step": 1879
     },
     {
       "epoch": 3.008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2958693504333496,
+      "learning_rate": 0.0005589,
+      "loss": 2.1657,
       "step": 1880
     },
     {
       "epoch": 3.0096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4474709033966064,
+      "learning_rate": 0.0005591999999999999,
+      "loss": 2.0949,
       "step": 1881
     },
     {
       "epoch": 3.0112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.083594560623169,
+      "learning_rate": 0.0005595,
+      "loss": 2.3073,
       "step": 1882
     },
     {
       "epoch": 3.0128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5751953125,
+      "learning_rate": 0.0005598,
+      "loss": 2.0699,
       "step": 1883
     },
     {
       "epoch": 3.0144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7937655448913574,
+      "learning_rate": 0.0005600999999999999,
+      "loss": 2.0185,
       "step": 1884
     },
     {
       "epoch": 3.016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5389955043792725,
+      "learning_rate": 0.0005604,
+      "loss": 1.7487,
       "step": 1885
     },
     {
       "epoch": 3.0176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.738939642906189,
+      "learning_rate": 0.0005606999999999999,
+      "loss": 1.9005,
       "step": 1886
     },
     {
       "epoch": 3.0192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3595213890075684,
+      "learning_rate": 0.000561,
+      "loss": 2.011,
       "step": 1887
     },
     {
       "epoch": 3.0208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.682566225528717,
+      "learning_rate": 0.0005612999999999999,
+      "loss": 1.8104,
       "step": 1888
     },
     {
       "epoch": 3.0224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8090096712112427,
+      "learning_rate": 0.0005616,
+      "loss": 1.7363,
       "step": 1889
     },
     {
       "epoch": 3.024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5234781503677368,
+      "learning_rate": 0.0005618999999999999,
+      "loss": 1.8555,
       "step": 1890
     },
     {
       "epoch": 3.0256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7015899419784546,
+      "learning_rate": 0.0005622,
+      "loss": 1.7636,
       "step": 1891
     },
     {
       "epoch": 3.0272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.202921390533447,
+      "learning_rate": 0.0005625,
+      "loss": 2.6288,
       "step": 1892
     },
     {
       "epoch": 3.0288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.232006549835205,
+      "learning_rate": 0.0005627999999999999,
+      "loss": 1.7678,
       "step": 1893
     },
     {
       "epoch": 3.0304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.543165683746338,
+      "learning_rate": 0.0005631,
+      "loss": 2.1546,
       "step": 1894
     },
     {
       "epoch": 3.032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.719556212425232,
+      "learning_rate": 0.0005633999999999999,
+      "loss": 1.7731,
       "step": 1895
     },
     {
       "epoch": 3.0336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8694783449172974,
+      "learning_rate": 0.0005637,
+      "loss": 1.7355,
       "step": 1896
     },
     {
       "epoch": 3.0352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0235390663146973,
+      "learning_rate": 0.0005639999999999999,
+      "loss": 1.9301,
       "step": 1897
     },
     {
       "epoch": 3.0368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5643758177757263,
+      "learning_rate": 0.0005643,
+      "loss": 2.0847,
       "step": 1898
     },
     {
       "epoch": 3.0384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.277967929840088,
+      "learning_rate": 0.0005646,
+      "loss": 2.1718,
       "step": 1899
     },
     {
       "epoch": 3.04,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 1900
-    },
-    {
-      "epoch": 3.04,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 160.4562,
-      "eval_samples_per_second": 19.544,
-      "eval_steps_per_second": 1.222,
-      "eval_wer": 1.0,
+      "grad_norm": 1.1477729082107544,
+      "learning_rate": 0.0005648999999999999,
+      "loss": 2.0403,
       "step": 1900
     },
     {
       "epoch": 3.0416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6605058908462524,
+      "learning_rate": 0.0005652,
+      "loss": 2.0047,
       "step": 1901
     },
     {
       "epoch": 3.0432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6357007622718811,
+      "learning_rate": 0.0005654999999999999,
+      "loss": 2.0975,
       "step": 1902
     },
     {
       "epoch": 3.0448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7702534198760986,
+      "learning_rate": 0.0005657999999999999,
+      "loss": 1.9427,
       "step": 1903
     },
     {
       "epoch": 3.0464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5849918127059937,
+      "learning_rate": 0.0005660999999999999,
+      "loss": 1.6914,
       "step": 1904
     },
     {
       "epoch": 3.048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6325346827507019,
+      "learning_rate": 0.0005663999999999999,
+      "loss": 1.7743,
       "step": 1905
     },
     {
       "epoch": 3.0496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7176706790924072,
+      "learning_rate": 0.0005667,
+      "loss": 1.9213,
       "step": 1906
     },
     {
       "epoch": 3.0512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0038143396377563,
+      "learning_rate": 0.0005669999999999999,
+      "loss": 1.9311,
       "step": 1907
     },
     {
       "epoch": 3.0528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6072099804878235,
+      "learning_rate": 0.0005673,
+      "loss": 2.0835,
       "step": 1908
     },
     {
       "epoch": 3.0544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0105469226837158,
+      "learning_rate": 0.0005675999999999999,
+      "loss": 1.8864,
       "step": 1909
     },
     {
       "epoch": 3.056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8080129027366638,
+      "learning_rate": 0.0005679,
+      "loss": 1.9892,
       "step": 1910
     },
     {
       "epoch": 3.0576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.45754092931747437,
+      "learning_rate": 0.0005681999999999999,
+      "loss": 1.9756,
       "step": 1911
     },
     {
       "epoch": 3.0592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.233243465423584,
+      "learning_rate": 0.0005685,
+      "loss": 1.8072,
       "step": 1912
     },
     {
       "epoch": 3.0608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.556039810180664,
+      "learning_rate": 0.0005688,
+      "loss": 1.7558,
       "step": 1913
     },
     {
       "epoch": 3.0624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.975345253944397,
+      "learning_rate": 0.0005691,
+      "loss": 1.6875,
       "step": 1914
     },
     {
       "epoch": 3.064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.621182918548584,
+      "learning_rate": 0.0005694,
+      "loss": 1.9418,
       "step": 1915
     },
     {
       "epoch": 3.0656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2835545539855957,
+      "learning_rate": 0.0005696999999999999,
+      "loss": 1.9206,
       "step": 1916
     },
     {
       "epoch": 3.0672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9312288761138916,
+      "learning_rate": 0.00057,
+      "loss": 2.0601,
       "step": 1917
     },
     {
       "epoch": 3.0688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1318893432617188,
+      "learning_rate": 0.0005702999999999999,
+      "loss": 2.1345,
       "step": 1918
     },
     {
       "epoch": 3.0704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6130589246749878,
+      "learning_rate": 0.0005705999999999999,
+      "loss": 2.1618,
       "step": 1919
     },
     {
       "epoch": 3.072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.576610803604126,
+      "learning_rate": 0.0005708999999999999,
+      "loss": 2.1284,
       "step": 1920
     },
     {
       "epoch": 3.0736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5201882719993591,
+      "learning_rate": 0.0005711999999999999,
+      "loss": 1.6506,
       "step": 1921
     },
     {
       "epoch": 3.0752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6475142240524292,
+      "learning_rate": 0.0005715,
+      "loss": 2.5413,
       "step": 1922
     },
     {
       "epoch": 3.0768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4755803346633911,
+      "learning_rate": 0.0005717999999999999,
+      "loss": 2.3249,
       "step": 1923
     },
     {
       "epoch": 3.0784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.319411516189575,
+      "learning_rate": 0.0005721,
+      "loss": 1.914,
       "step": 1924
     },
     {
       "epoch": 3.08,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6974052786827087,
+      "learning_rate": 0.0005723999999999999,
+      "loss": 2.448,
       "step": 1925
     },
     {
       "epoch": 3.0816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.256194114685059,
+      "learning_rate": 0.0005727,
+      "loss": 2.5141,
       "step": 1926
     },
     {
       "epoch": 3.0832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4125750064849854,
+      "learning_rate": 0.0005729999999999999,
+      "loss": 2.3816,
       "step": 1927
     },
     {
       "epoch": 3.0848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0927053689956665,
+      "learning_rate": 0.0005733,
+      "loss": 2.0794,
       "step": 1928
     },
     {
       "epoch": 3.0864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9806353449821472,
+      "learning_rate": 0.0005736,
+      "loss": 2.0871,
       "step": 1929
     },
     {
       "epoch": 3.088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8955163955688477,
+      "learning_rate": 0.0005738999999999999,
+      "loss": 2.462,
       "step": 1930
     },
     {
       "epoch": 3.0896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9962303638458252,
+      "learning_rate": 0.0005742,
+      "loss": 2.1367,
       "step": 1931
     },
     {
       "epoch": 3.0912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.6139843463897705,
+      "learning_rate": 0.0005744999999999999,
+      "loss": 2.0657,
       "step": 1932
     },
     {
       "epoch": 3.0928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9958500862121582,
+      "learning_rate": 0.0005747999999999999,
+      "loss": 2.1125,
       "step": 1933
     },
     {
       "epoch": 3.0944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8471503257751465,
+      "learning_rate": 0.0005750999999999999,
+      "loss": 1.7954,
       "step": 1934
     },
     {
       "epoch": 3.096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0833854675292969,
+      "learning_rate": 0.0005753999999999999,
+      "loss": 2.1097,
       "step": 1935
     },
     {
       "epoch": 3.0976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8012189865112305,
+      "learning_rate": 0.0005757,
+      "loss": 1.9832,
       "step": 1936
     },
     {
       "epoch": 3.0992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5345687866210938,
+      "learning_rate": 0.0005759999999999999,
+      "loss": 1.7323,
       "step": 1937
     },
     {
       "epoch": 3.1008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6692140102386475,
+      "learning_rate": 0.0005763,
+      "loss": 2.145,
       "step": 1938
     },
     {
       "epoch": 3.1024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1121718883514404,
+      "learning_rate": 0.0005765999999999999,
+      "loss": 1.8593,
       "step": 1939
     },
     {
       "epoch": 3.104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7572036981582642,
+      "learning_rate": 0.0005769,
+      "loss": 1.7605,
       "step": 1940
     },
     {
       "epoch": 3.1056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4940950274467468,
+      "learning_rate": 0.0005771999999999999,
+      "loss": 1.7811,
       "step": 1941
     },
     {
       "epoch": 3.1072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4116694927215576,
+      "learning_rate": 0.0005775,
+      "loss": 1.7468,
       "step": 1942
     },
     {
       "epoch": 3.1088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1187795400619507,
+      "learning_rate": 0.0005778,
+      "loss": 2.1021,
       "step": 1943
     },
     {
       "epoch": 3.1104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.035007953643799,
+      "learning_rate": 0.0005781,
+      "loss": 2.0287,
       "step": 1944
     },
     {
       "epoch": 3.112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9366320371627808,
+      "learning_rate": 0.0005784,
+      "loss": 1.8002,
       "step": 1945
     },
     {
       "epoch": 3.1136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9380377531051636,
+      "learning_rate": 0.0005786999999999999,
+      "loss": 1.9192,
       "step": 1946
     },
     {
       "epoch": 3.1152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6677077412605286,
+      "learning_rate": 0.000579,
+      "loss": 1.8212,
       "step": 1947
     },
     {
       "epoch": 3.1168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8764153718948364,
+      "learning_rate": 0.0005792999999999999,
+      "loss": 1.9796,
       "step": 1948
     },
     {
       "epoch": 3.1184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.44954901933670044,
+      "learning_rate": 0.0005795999999999999,
+      "loss": 1.7954,
       "step": 1949
     },
     {
       "epoch": 3.12,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.038944125175476,
+      "learning_rate": 0.0005799,
+      "loss": 2.1065,
       "step": 1950
     },
     {
       "epoch": 3.1216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.420203924179077,
+      "learning_rate": 0.0005801999999999999,
+      "loss": 1.8525,
       "step": 1951
     },
     {
       "epoch": 3.1232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3937735557556152,
+      "learning_rate": 0.0005805,
+      "loss": 1.8255,
       "step": 1952
     },
     {
       "epoch": 3.1248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7212362289428711,
+      "learning_rate": 0.0005807999999999999,
+      "loss": 1.8263,
       "step": 1953
     },
     {
       "epoch": 3.1264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8853600025177002,
+      "learning_rate": 0.0005811,
+      "loss": 1.8761,
       "step": 1954
     },
     {
       "epoch": 3.128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4678289890289307,
+      "learning_rate": 0.0005813999999999999,
+      "loss": 2.0627,
       "step": 1955
     },
     {
       "epoch": 3.1296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1042805910110474,
+      "learning_rate": 0.0005817,
+      "loss": 1.8495,
       "step": 1956
     },
     {
       "epoch": 3.1312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6540318131446838,
+      "learning_rate": 0.0005819999999999999,
+      "loss": 1.8383,
       "step": 1957
     },
     {
       "epoch": 3.1328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8567636013031006,
+      "learning_rate": 0.0005823,
+      "loss": 1.8997,
       "step": 1958
     },
     {
       "epoch": 3.1344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.660646677017212,
+      "learning_rate": 0.0005826,
+      "loss": 1.8058,
       "step": 1959
     },
     {
       "epoch": 3.136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4510436058044434,
+      "learning_rate": 0.0005829,
+      "loss": 1.9501,
       "step": 1960
     },
     {
       "epoch": 3.1376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.678549289703369,
+      "learning_rate": 0.0005832,
+      "loss": 2.2273,
       "step": 1961
     },
     {
       "epoch": 3.1391999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9886896014213562,
+      "learning_rate": 0.0005834999999999999,
+      "loss": 1.8041,
       "step": 1962
     },
     {
       "epoch": 3.1408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7258390188217163,
+      "learning_rate": 0.0005838,
+      "loss": 1.9527,
       "step": 1963
     },
     {
       "epoch": 3.1424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1479647159576416,
+      "learning_rate": 0.0005840999999999999,
+      "loss": 2.2193,
       "step": 1964
     },
     {
       "epoch": 3.144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9991073608398438,
+      "learning_rate": 0.0005843999999999999,
+      "loss": 2.1787,
       "step": 1965
     },
     {
       "epoch": 3.1456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5090608596801758,
+      "learning_rate": 0.0005847,
+      "loss": 1.9899,
       "step": 1966
     },
     {
       "epoch": 3.1471999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6410496830940247,
+      "learning_rate": 0.0005849999999999999,
+      "loss": 1.7924,
       "step": 1967
     },
     {
       "epoch": 3.1488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2429163455963135,
+      "learning_rate": 0.0005853,
+      "loss": 2.1826,
       "step": 1968
     },
     {
       "epoch": 3.1504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9956017136573792,
+      "learning_rate": 0.0005855999999999999,
+      "loss": 1.994,
       "step": 1969
     },
     {
       "epoch": 3.152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.732032299041748,
+      "learning_rate": 0.0005859,
+      "loss": 2.0436,
       "step": 1970
     },
     {
       "epoch": 3.1536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.820035219192505,
+      "learning_rate": 0.0005861999999999999,
+      "loss": 1.8376,
       "step": 1971
     },
     {
       "epoch": 3.1552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5163549184799194,
+      "learning_rate": 0.0005865,
+      "loss": 2.1958,
       "step": 1972
     },
     {
       "epoch": 3.1568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5381282567977905,
+      "learning_rate": 0.0005868,
+      "loss": 2.3952,
       "step": 1973
     },
     {
       "epoch": 3.1584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8239006400108337,
+      "learning_rate": 0.0005871,
+      "loss": 2.3353,
       "step": 1974
     },
     {
       "epoch": 3.16,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8696528673171997,
+      "learning_rate": 0.0005874,
+      "loss": 2.0183,
       "step": 1975
     },
     {
       "epoch": 3.1616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.004875183105469,
+      "learning_rate": 0.0005876999999999999,
+      "loss": 2.9254,
       "step": 1976
     },
     {
       "epoch": 3.1632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3852179050445557,
+      "learning_rate": 0.000588,
+      "loss": 2.2966,
       "step": 1977
     },
     {
       "epoch": 3.1648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2804005146026611,
+      "learning_rate": 0.0005882999999999999,
+      "loss": 2.3582,
       "step": 1978
     },
     {
       "epoch": 3.1664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.041597366333008,
+      "learning_rate": 0.0005885999999999999,
+      "loss": 2.702,
       "step": 1979
     },
     {
       "epoch": 3.168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.890049934387207,
+      "learning_rate": 0.0005889,
+      "loss": 2.0844,
       "step": 1980
     },
     {
       "epoch": 3.1696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.977160692214966,
+      "learning_rate": 0.0005891999999999999,
+      "loss": 2.2058,
       "step": 1981
     },
     {
       "epoch": 3.1712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.654447078704834,
+      "learning_rate": 0.0005895,
+      "loss": 2.1782,
       "step": 1982
     },
     {
       "epoch": 3.1728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9558517932891846,
+      "learning_rate": 0.0005897999999999999,
+      "loss": 2.2067,
       "step": 1983
     },
     {
       "epoch": 3.1744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9267162084579468,
+      "learning_rate": 0.0005901,
+      "loss": 1.9776,
       "step": 1984
     },
     {
       "epoch": 3.176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7199661731719971,
+      "learning_rate": 0.0005903999999999999,
+      "loss": 2.1208,
       "step": 1985
     },
     {
       "epoch": 3.1776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.00618052482605,
+      "learning_rate": 0.0005907,
+      "loss": 1.991,
       "step": 1986
     },
     {
       "epoch": 3.1792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5676202774047852,
+      "learning_rate": 0.0005909999999999999,
+      "loss": 2.0481,
       "step": 1987
     },
     {
       "epoch": 3.1808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.522994041442871,
+      "learning_rate": 0.0005913,
+      "loss": 2.1883,
       "step": 1988
     },
     {
       "epoch": 3.1824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1395773887634277,
+      "learning_rate": 0.0005916,
+      "loss": 1.7289,
       "step": 1989
     },
     {
       "epoch": 3.184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1741175651550293,
+      "learning_rate": 0.0005919,
+      "loss": 1.8146,
       "step": 1990
     },
     {
       "epoch": 3.1856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9648697972297668,
+      "learning_rate": 0.0005922,
+      "loss": 1.9346,
       "step": 1991
     },
     {
       "epoch": 3.1872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.453245759010315,
+      "learning_rate": 0.0005924999999999999,
+      "loss": 1.887,
       "step": 1992
     },
     {
       "epoch": 3.1888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3214921951293945,
+      "learning_rate": 0.0005928,
+      "loss": 1.9685,
       "step": 1993
     },
     {
       "epoch": 3.1904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6422308683395386,
+      "learning_rate": 0.0005930999999999999,
+      "loss": 1.7544,
       "step": 1994
     },
     {
       "epoch": 3.192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8797899484634399,
+      "learning_rate": 0.0005933999999999999,
+      "loss": 2.1596,
       "step": 1995
     },
     {
       "epoch": 3.1936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5090241432189941,
+      "learning_rate": 0.0005937,
+      "loss": 1.697,
       "step": 1996
     },
     {
       "epoch": 3.1952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.829024612903595,
+      "learning_rate": 0.0005939999999999999,
+      "loss": 2.0842,
       "step": 1997
     },
     {
       "epoch": 3.1968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7225773930549622,
+      "learning_rate": 0.0005943,
+      "loss": 1.8067,
       "step": 1998
     },
     {
       "epoch": 3.1984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5545061230659485,
+      "learning_rate": 0.0005945999999999999,
+      "loss": 1.874,
       "step": 1999
     },
     {
       "epoch": 3.2,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.47109493613243103,
+      "learning_rate": 0.0005949,
+      "loss": 1.8352,
       "step": 2000
     },
     {
       "epoch": 3.2,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 159.6533,
-      "eval_samples_per_second": 19.643,
-      "eval_steps_per_second": 1.228,
-      "eval_wer": 1.0,
+      "eval_cer": 0.472372144935624,
+      "eval_loss": 2.1490743160247803,
+      "eval_runtime": 158.1752,
+      "eval_samples_per_second": 19.826,
+      "eval_steps_per_second": 1.239,
+      "eval_wer": 0.7894725935136255,
       "step": 2000
     },
     {
       "epoch": 3.2016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.672227382659912,
+      "learning_rate": 0.0005951999999999999,
+      "loss": 1.8896,
       "step": 2001
     },
     {
       "epoch": 3.2032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8217174410820007,
+      "learning_rate": 0.0005955,
+      "loss": 1.7263,
       "step": 2002
     },
     {
       "epoch": 3.2048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7968158721923828,
+      "learning_rate": 0.0005958,
+      "loss": 2.1543,
       "step": 2003
     },
     {
       "epoch": 3.2064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.697176456451416,
+      "learning_rate": 0.0005961,
+      "loss": 1.9787,
       "step": 2004
     },
     {
       "epoch": 3.208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1341164112091064,
+      "learning_rate": 0.0005964,
+      "loss": 1.9606,
       "step": 2005
     },
     {
       "epoch": 3.2096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5551044940948486,
+      "learning_rate": 0.0005967,
+      "loss": 1.7492,
       "step": 2006
     },
     {
       "epoch": 3.2112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8107208013534546,
+      "learning_rate": 0.000597,
+      "loss": 1.936,
       "step": 2007
     },
     {
       "epoch": 3.2128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5374244451522827,
+      "learning_rate": 0.0005972999999999999,
+      "loss": 2.0281,
       "step": 2008
     },
     {
       "epoch": 3.2144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6073170900344849,
+      "learning_rate": 0.0005976,
+      "loss": 1.9526,
       "step": 2009
     },
     {
       "epoch": 3.216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5787049531936646,
+      "learning_rate": 0.0005979,
+      "loss": 1.7531,
       "step": 2010
     },
     {
       "epoch": 3.2176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9718129634857178,
+      "learning_rate": 0.0005981999999999999,
+      "loss": 1.9827,
       "step": 2011
     },
     {
       "epoch": 3.2192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6538378596305847,
+      "learning_rate": 0.0005985,
+      "loss": 1.9704,
       "step": 2012
     },
     {
       "epoch": 3.2208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7849467992782593,
+      "learning_rate": 0.0005987999999999999,
+      "loss": 1.9564,
       "step": 2013
     },
     {
       "epoch": 3.2224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7062731981277466,
+      "learning_rate": 0.0005991,
+      "loss": 1.8622,
       "step": 2014
     },
     {
       "epoch": 3.224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.715021014213562,
+      "learning_rate": 0.0005993999999999999,
+      "loss": 2.0217,
       "step": 2015
     },
     {
       "epoch": 3.2256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1211111545562744,
+      "learning_rate": 0.0005997,
+      "loss": 2.2437,
       "step": 2016
     },
     {
       "epoch": 3.2272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8204267621040344,
+      "learning_rate": 0.0006,
+      "loss": 1.9696,
       "step": 2017
     },
     {
       "epoch": 3.2288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9571385383605957,
+      "learning_rate": 0.00059985,
+      "loss": 1.8388,
       "step": 2018
     },
     {
       "epoch": 3.2304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6814630627632141,
+      "learning_rate": 0.0005997,
+      "loss": 1.8718,
       "step": 2019
     },
     {
       "epoch": 3.232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.779271125793457,
+      "learning_rate": 0.0005995499999999999,
+      "loss": 2.0305,
       "step": 2020
     },
     {
       "epoch": 3.2336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7469867467880249,
+      "learning_rate": 0.0005993999999999999,
+      "loss": 1.7589,
       "step": 2021
     },
     {
       "epoch": 3.2352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8041962385177612,
+      "learning_rate": 0.00059925,
+      "loss": 1.7746,
       "step": 2022
     },
     {
       "epoch": 3.2368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6314986944198608,
+      "learning_rate": 0.0005991,
+      "loss": 2.4096,
       "step": 2023
     },
     {
       "epoch": 3.2384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6365277767181396,
+      "learning_rate": 0.0005989499999999999,
+      "loss": 2.1849,
       "step": 2024
     },
     {
       "epoch": 3.24,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8172062039375305,
+      "learning_rate": 0.0005987999999999999,
+      "loss": 2.0949,
       "step": 2025
     },
     {
       "epoch": 3.2416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4743893146514893,
+      "learning_rate": 0.0005986499999999999,
+      "loss": 2.3107,
       "step": 2026
     },
     {
       "epoch": 3.2432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.591100215911865,
+      "learning_rate": 0.0005985,
+      "loss": 2.4814,
       "step": 2027
     },
     {
       "epoch": 3.2448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.602311134338379,
+      "learning_rate": 0.0005983499999999999,
+      "loss": 2.8672,
       "step": 2028
     },
     {
       "epoch": 3.2464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2622711658477783,
+      "learning_rate": 0.0005981999999999999,
+      "loss": 2.4354,
       "step": 2029
     },
     {
       "epoch": 3.248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.489107131958008,
+      "learning_rate": 0.0005980499999999999,
+      "loss": 2.2468,
       "step": 2030
     },
     {
       "epoch": 3.2496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.640115737915039,
+      "learning_rate": 0.0005979,
+      "loss": 2.2934,
       "step": 2031
     },
     {
       "epoch": 3.2512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.885188579559326,
+      "learning_rate": 0.00059775,
+      "loss": 2.5731,
       "step": 2032
     },
     {
       "epoch": 3.2528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.099841117858887,
+      "learning_rate": 0.0005976,
+      "loss": 2.323,
       "step": 2033
     },
     {
       "epoch": 3.2544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.89034366607666,
+      "learning_rate": 0.0005974499999999999,
+      "loss": 2.3686,
       "step": 2034
     },
     {
       "epoch": 3.2560000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.048368453979492,
+      "learning_rate": 0.0005972999999999999,
+      "loss": 2.0215,
       "step": 2035
     },
     {
       "epoch": 3.2576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.314151287078857,
+      "learning_rate": 0.00059715,
+      "loss": 2.0318,
       "step": 2036
     },
     {
       "epoch": 3.2592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.657911777496338,
+      "learning_rate": 0.000597,
+      "loss": 2.1174,
       "step": 2037
     },
     {
       "epoch": 3.2608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7400920391082764,
+      "learning_rate": 0.00059685,
+      "loss": 2.2229,
       "step": 2038
     },
     {
       "epoch": 3.2624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.803786039352417,
+      "learning_rate": 0.0005967,
+      "loss": 1.8763,
       "step": 2039
     },
     {
       "epoch": 3.2640000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4826065301895142,
+      "learning_rate": 0.0005965499999999999,
+      "loss": 1.9995,
       "step": 2040
     },
     {
       "epoch": 3.2656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1381956338882446,
+      "learning_rate": 0.0005964,
+      "loss": 1.7465,
       "step": 2041
     },
     {
       "epoch": 3.2672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0622131824493408,
+      "learning_rate": 0.00059625,
+      "loss": 2.011,
       "step": 2042
     },
     {
       "epoch": 3.2688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2198865413665771,
+      "learning_rate": 0.0005961,
+      "loss": 1.9016,
       "step": 2043
     },
     {
       "epoch": 3.2704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8227574825286865,
+      "learning_rate": 0.0005959499999999999,
+      "loss": 1.9384,
       "step": 2044
     },
     {
       "epoch": 3.2720000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4436991214752197,
+      "learning_rate": 0.0005958,
+      "loss": 1.8939,
       "step": 2045
     },
     {
       "epoch": 3.2736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6856980323791504,
+      "learning_rate": 0.00059565,
+      "loss": 1.9194,
       "step": 2046
     },
     {
       "epoch": 3.2752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3934940099716187,
+      "learning_rate": 0.0005955,
+      "loss": 1.8451,
       "step": 2047
     },
     {
       "epoch": 3.2768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9621049165725708,
+      "learning_rate": 0.0005953499999999999,
+      "loss": 2.1891,
       "step": 2048
     },
     {
       "epoch": 3.2784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1755985021591187,
+      "learning_rate": 0.0005951999999999999,
+      "loss": 1.9826,
       "step": 2049
     },
     {
       "epoch": 3.2800000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0933473110198975,
+      "learning_rate": 0.00059505,
+      "loss": 1.7576,
       "step": 2050
     },
     {
       "epoch": 3.2816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2009758949279785,
+      "learning_rate": 0.0005949,
+      "loss": 1.9181,
       "step": 2051
     },
     {
       "epoch": 3.2832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6885703206062317,
+      "learning_rate": 0.0005947499999999999,
+      "loss": 1.6412,
       "step": 2052
     },
     {
       "epoch": 3.2848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3572481870651245,
+      "learning_rate": 0.0005945999999999999,
+      "loss": 2.0448,
       "step": 2053
     },
     {
       "epoch": 3.2864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.438791036605835,
+      "learning_rate": 0.00059445,
+      "loss": 1.9378,
       "step": 2054
     },
     {
       "epoch": 3.288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.896614670753479,
+      "learning_rate": 0.0005943,
+      "loss": 1.8265,
       "step": 2055
     },
     {
       "epoch": 3.2896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.369891405105591,
+      "learning_rate": 0.0005941499999999999,
+      "loss": 1.9555,
       "step": 2056
     },
     {
       "epoch": 3.2912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7984668016433716,
+      "learning_rate": 0.0005939999999999999,
+      "loss": 1.8772,
       "step": 2057
     },
     {
       "epoch": 3.2928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.576426088809967,
+      "learning_rate": 0.0005938499999999999,
+      "loss": 1.7301,
       "step": 2058
     },
     {
       "epoch": 3.2944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9307464361190796,
+      "learning_rate": 0.0005937,
+      "loss": 1.9711,
       "step": 2059
     },
     {
       "epoch": 3.296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5098490715026855,
+      "learning_rate": 0.0005935499999999999,
+      "loss": 1.9865,
       "step": 2060
     },
     {
       "epoch": 3.2976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9491452574729919,
+      "learning_rate": 0.0005933999999999999,
+      "loss": 2.0115,
       "step": 2061
     },
     {
       "epoch": 3.2992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.528455913066864,
+      "learning_rate": 0.0005932499999999999,
+      "loss": 1.8317,
       "step": 2062
     },
     {
       "epoch": 3.3008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3873507976531982,
+      "learning_rate": 0.0005930999999999999,
+      "loss": 2.2621,
       "step": 2063
     },
     {
       "epoch": 3.3024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1286377906799316,
+      "learning_rate": 0.00059295,
+      "loss": 2.0241,
       "step": 2064
     },
     {
       "epoch": 3.304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4034842252731323,
+      "learning_rate": 0.0005928,
+      "loss": 1.9359,
       "step": 2065
     },
     {
       "epoch": 3.3056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8125107884407043,
+      "learning_rate": 0.0005926499999999999,
+      "loss": 2.0162,
       "step": 2066
     },
     {
       "epoch": 3.3072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2552961111068726,
+      "learning_rate": 0.0005924999999999999,
+      "loss": 2.1944,
       "step": 2067
     },
     {
       "epoch": 3.3088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9320792555809021,
+      "learning_rate": 0.00059235,
+      "loss": 2.109,
       "step": 2068
     },
     {
       "epoch": 3.3104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9960517883300781,
+      "learning_rate": 0.0005922,
+      "loss": 2.0276,
       "step": 2069
     },
     {
       "epoch": 3.312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9736140370368958,
+      "learning_rate": 0.00059205,
+      "loss": 2.1911,
       "step": 2070
     },
     {
       "epoch": 3.3136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5229682922363281,
+      "learning_rate": 0.0005919,
+      "loss": 2.1335,
       "step": 2071
     },
     {
       "epoch": 3.3152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.261260747909546,
+      "learning_rate": 0.0005917499999999999,
+      "loss": 2.0071,
       "step": 2072
     },
     {
       "epoch": 3.3168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7125620245933533,
+      "learning_rate": 0.0005916,
+      "loss": 2.1277,
       "step": 2073
     },
     {
       "epoch": 3.3184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.672423839569092,
+      "learning_rate": 0.00059145,
+      "loss": 2.2381,
       "step": 2074
     },
     {
       "epoch": 3.32,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00059145,
+      "loss": 2.606,
       "step": 2075
     },
     {
       "epoch": 3.3216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 12.852534294128418,
+      "learning_rate": 0.0005913,
+      "loss": 2.6536,
       "step": 2076
     },
     {
       "epoch": 3.3232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6311789751052856,
+      "learning_rate": 0.0005911499999999999,
+      "loss": 2.2505,
       "step": 2077
     },
     {
       "epoch": 3.3247999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5713582038879395,
+      "learning_rate": 0.0005909999999999999,
+      "loss": 2.1638,
       "step": 2078
     },
     {
       "epoch": 3.3264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8965359926223755,
+      "learning_rate": 0.00059085,
+      "loss": 2.3546,
       "step": 2079
     },
     {
       "epoch": 3.328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8224315643310547,
+      "learning_rate": 0.0005907,
+      "loss": 2.1339,
       "step": 2080
     },
     {
       "epoch": 3.3296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.236239433288574,
+      "learning_rate": 0.0005905499999999999,
+      "loss": 2.1391,
       "step": 2081
     },
     {
       "epoch": 3.3312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.328113555908203,
+      "learning_rate": 0.0005903999999999999,
+      "loss": 1.9743,
       "step": 2082
     },
     {
       "epoch": 3.3327999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0247113704681396,
+      "learning_rate": 0.00059025,
+      "loss": 2.1961,
       "step": 2083
     },
     {
       "epoch": 3.3344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5589113235473633,
+      "learning_rate": 0.0005901,
+      "loss": 1.904,
       "step": 2084
     },
     {
       "epoch": 3.336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7790500521659851,
+      "learning_rate": 0.0005899499999999999,
+      "loss": 2.1242,
       "step": 2085
     },
     {
       "epoch": 3.3376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.423574924468994,
+      "learning_rate": 0.0005897999999999999,
+      "loss": 2.1614,
       "step": 2086
     },
     {
       "epoch": 3.3392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7823060750961304,
+      "learning_rate": 0.0005896499999999999,
+      "loss": 2.0141,
       "step": 2087
     },
     {
       "epoch": 3.3407999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7111868858337402,
+      "learning_rate": 0.0005895,
+      "loss": 2.097,
       "step": 2088
     },
     {
       "epoch": 3.3424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6286764144897461,
+      "learning_rate": 0.0005893499999999999,
+      "loss": 1.8148,
       "step": 2089
     },
     {
       "epoch": 3.344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0036760568618774,
+      "learning_rate": 0.0005891999999999999,
+      "loss": 1.7824,
       "step": 2090
     },
     {
       "epoch": 3.3456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.625751256942749,
+      "learning_rate": 0.0005890499999999999,
+      "loss": 2.0269,
       "step": 2091
     },
     {
       "epoch": 3.3472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7607942223548889,
+      "learning_rate": 0.0005889,
+      "loss": 1.7751,
       "step": 2092
     },
     {
       "epoch": 3.3487999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3163102865219116,
+      "learning_rate": 0.00058875,
+      "loss": 1.9094,
       "step": 2093
     },
     {
       "epoch": 3.3504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.412693738937378,
+      "learning_rate": 0.0005885999999999999,
+      "loss": 1.7847,
       "step": 2094
     },
     {
       "epoch": 3.352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.028609275817871,
+      "learning_rate": 0.0005884499999999999,
+      "loss": 1.9335,
       "step": 2095
     },
     {
       "epoch": 3.3536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.106916904449463,
+      "learning_rate": 0.0005882999999999999,
+      "loss": 1.9819,
       "step": 2096
     },
     {
       "epoch": 3.3552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9769391417503357,
+      "learning_rate": 0.00058815,
+      "loss": 2.178,
       "step": 2097
     },
     {
       "epoch": 3.3568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7147427797317505,
+      "learning_rate": 0.000588,
+      "loss": 1.7971,
       "step": 2098
     },
     {
       "epoch": 3.3584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.773476779460907,
+      "learning_rate": 0.00058785,
+      "loss": 1.7747,
       "step": 2099
     },
     {
       "epoch": 3.36,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2100
-    },
-    {
-      "epoch": 3.36,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 161.4742,
-      "eval_samples_per_second": 19.421,
-      "eval_steps_per_second": 1.214,
-      "eval_wer": 1.0,
+      "grad_norm": 0.7806805372238159,
+      "learning_rate": 0.0005876999999999999,
+      "loss": 1.9335,
       "step": 2100
     },
     {
       "epoch": 3.3616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0920946598052979,
+      "learning_rate": 0.0005875499999999999,
+      "loss": 1.8085,
       "step": 2101
     },
     {
       "epoch": 3.3632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8656213283538818,
+      "learning_rate": 0.0005874,
+      "loss": 2.0736,
       "step": 2102
     },
     {
       "epoch": 3.3648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.247507095336914,
+      "learning_rate": 0.00058725,
+      "loss": 2.0338,
       "step": 2103
     },
     {
       "epoch": 3.3664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6680485010147095,
+      "learning_rate": 0.0005871,
+      "loss": 1.9589,
       "step": 2104
     },
     {
       "epoch": 3.368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4975533485412598,
+      "learning_rate": 0.0005869499999999999,
+      "loss": 1.9832,
       "step": 2105
     },
     {
       "epoch": 3.3696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8459756374359131,
+      "learning_rate": 0.0005868,
+      "loss": 1.8083,
       "step": 2106
     },
     {
       "epoch": 3.3712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.097440242767334,
+      "learning_rate": 0.00058665,
+      "loss": 1.661,
       "step": 2107
     },
     {
       "epoch": 3.3728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0254461765289307,
+      "learning_rate": 0.0005865,
+      "loss": 1.9348,
       "step": 2108
     },
     {
       "epoch": 3.3744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9976134896278381,
+      "learning_rate": 0.0005863499999999999,
+      "loss": 1.7156,
       "step": 2109
     },
     {
       "epoch": 3.376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9642464518547058,
+      "learning_rate": 0.0005861999999999999,
+      "loss": 1.8166,
       "step": 2110
     },
     {
       "epoch": 3.3776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2344017028808594,
+      "learning_rate": 0.00058605,
+      "loss": 2.4248,
       "step": 2111
     },
     {
       "epoch": 3.3792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6256465315818787,
+      "learning_rate": 0.0005859,
+      "loss": 2.1543,
       "step": 2112
     },
     {
       "epoch": 3.3808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9947925806045532,
+      "learning_rate": 0.0005857499999999999,
+      "loss": 1.9461,
       "step": 2113
     },
     {
       "epoch": 3.3824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0438300371170044,
+      "learning_rate": 0.0005855999999999999,
+      "loss": 2.2066,
       "step": 2114
     },
     {
       "epoch": 3.384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5882972478866577,
+      "learning_rate": 0.00058545,
+      "loss": 2.1843,
       "step": 2115
     },
     {
       "epoch": 3.3856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0569475889205933,
+      "learning_rate": 0.0005853,
+      "loss": 1.9829,
       "step": 2116
     },
     {
       "epoch": 3.3872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5413305163383484,
+      "learning_rate": 0.0005851499999999999,
+      "loss": 2.0784,
       "step": 2117
     },
     {
       "epoch": 3.3888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.626608669757843,
+      "learning_rate": 0.0005849999999999999,
+      "loss": 1.7897,
       "step": 2118
     },
     {
       "epoch": 3.3904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6285375356674194,
+      "learning_rate": 0.0005848499999999999,
+      "loss": 1.9263,
       "step": 2119
     },
     {
       "epoch": 3.392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4336178302764893,
+      "learning_rate": 0.0005847,
+      "loss": 1.9986,
       "step": 2120
     },
     {
       "epoch": 3.3936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0149215459823608,
+      "learning_rate": 0.0005845499999999999,
+      "loss": 1.8295,
       "step": 2121
     },
     {
       "epoch": 3.3952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.226423740386963,
+      "learning_rate": 0.0005843999999999999,
+      "loss": 2.2674,
       "step": 2122
     },
     {
       "epoch": 3.3968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9696506261825562,
+      "learning_rate": 0.0005842499999999999,
+      "loss": 1.9007,
       "step": 2123
     },
     {
       "epoch": 3.3984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1912128925323486,
+      "learning_rate": 0.0005840999999999999,
+      "loss": 2.5396,
       "step": 2124
     },
     {
       "epoch": 3.4,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.123094081878662,
+      "learning_rate": 0.00058395,
+      "loss": 2.2945,
       "step": 2125
     },
     {
       "epoch": 3.4016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7035903930664062,
+      "learning_rate": 0.0005838,
+      "loss": 2.487,
       "step": 2126
     },
     {
       "epoch": 3.4032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1892151832580566,
+      "learning_rate": 0.0005836499999999999,
+      "loss": 2.3695,
       "step": 2127
     },
     {
       "epoch": 3.4048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2689785957336426,
+      "learning_rate": 0.0005834999999999999,
+      "loss": 2.2589,
       "step": 2128
     },
     {
       "epoch": 3.4064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1017853021621704,
+      "learning_rate": 0.00058335,
+      "loss": 2.3216,
       "step": 2129
     },
     {
       "epoch": 3.408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3482823371887207,
+      "learning_rate": 0.0005832,
+      "loss": 2.0859,
       "step": 2130
     },
     {
       "epoch": 3.4096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.300066947937012,
+      "learning_rate": 0.00058305,
+      "loss": 2.5814,
       "step": 2131
     },
     {
       "epoch": 3.4112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3666080236434937,
+      "learning_rate": 0.0005829,
+      "loss": 2.4822,
       "step": 2132
     },
     {
       "epoch": 3.4128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.999903678894043,
+      "learning_rate": 0.0005827499999999999,
+      "loss": 1.9601,
       "step": 2133
     },
     {
       "epoch": 3.4144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.403241157531738,
+      "learning_rate": 0.0005826,
+      "loss": 2.1051,
       "step": 2134
     },
     {
       "epoch": 3.416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.511754989624023,
+      "learning_rate": 0.00058245,
+      "loss": 2.1249,
       "step": 2135
     },
     {
       "epoch": 3.4176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8616156578063965,
+      "learning_rate": 0.0005823,
+      "loss": 2.02,
       "step": 2136
     },
     {
       "epoch": 3.4192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.203505516052246,
+      "learning_rate": 0.0005821499999999999,
+      "loss": 2.0634,
       "step": 2137
     },
     {
       "epoch": 3.4208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.6515045166015625,
+      "learning_rate": 0.0005819999999999999,
+      "loss": 1.9337,
       "step": 2138
     },
     {
       "epoch": 3.4224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.657661199569702,
+      "learning_rate": 0.00058185,
+      "loss": 1.8973,
       "step": 2139
     },
     {
       "epoch": 3.424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9265851974487305,
+      "learning_rate": 0.0005817,
+      "loss": 2.1529,
       "step": 2140
     },
     {
       "epoch": 3.4256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6995819211006165,
+      "learning_rate": 0.0005815499999999999,
+      "loss": 1.9286,
       "step": 2141
     },
     {
       "epoch": 3.4272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5046209096908569,
+      "learning_rate": 0.0005813999999999999,
+      "loss": 1.8917,
       "step": 2142
     },
     {
       "epoch": 3.4288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5500861406326294,
+      "learning_rate": 0.00058125,
+      "loss": 1.771,
       "step": 2143
     },
     {
       "epoch": 3.4304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7403568029403687,
+      "learning_rate": 0.0005811,
+      "loss": 1.9863,
       "step": 2144
     },
     {
       "epoch": 3.432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9957071542739868,
+      "learning_rate": 0.00058095,
+      "loss": 2.0033,
       "step": 2145
     },
     {
       "epoch": 3.4336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2035460472106934,
+      "learning_rate": 0.0005807999999999999,
+      "loss": 1.708,
       "step": 2146
     },
     {
       "epoch": 3.4352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9599239826202393,
+      "learning_rate": 0.0005806499999999999,
+      "loss": 2.2345,
       "step": 2147
     },
     {
       "epoch": 3.4368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.531463384628296,
+      "learning_rate": 0.0005805,
+      "loss": 2.0405,
       "step": 2148
     },
     {
       "epoch": 3.4384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0532867908477783,
+      "learning_rate": 0.00058035,
+      "loss": 1.824,
       "step": 2149
     },
     {
       "epoch": 3.44,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9876016974449158,
+      "learning_rate": 0.0005801999999999999,
+      "loss": 1.8832,
       "step": 2150
     },
     {
       "epoch": 3.4416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5361031293869019,
+      "learning_rate": 0.0005800499999999999,
+      "loss": 1.9323,
       "step": 2151
     },
     {
       "epoch": 3.4432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4722892045974731,
+      "learning_rate": 0.0005799,
+      "loss": 1.72,
       "step": 2152
     },
     {
       "epoch": 3.4448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8053250312805176,
+      "learning_rate": 0.00057975,
+      "loss": 2.1203,
       "step": 2153
     },
     {
       "epoch": 3.4464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2486746311187744,
+      "learning_rate": 0.0005795999999999999,
+      "loss": 1.6535,
       "step": 2154
     },
     {
       "epoch": 3.448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.140870451927185,
+      "learning_rate": 0.0005794499999999999,
+      "loss": 2.0749,
       "step": 2155
     },
     {
       "epoch": 3.4496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8704653978347778,
+      "learning_rate": 0.0005792999999999999,
+      "loss": 2.0133,
       "step": 2156
     },
     {
       "epoch": 3.4512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.039218544960022,
+      "learning_rate": 0.00057915,
+      "loss": 1.8325,
       "step": 2157
     },
     {
       "epoch": 3.4528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5828589200973511,
+      "learning_rate": 0.000579,
+      "loss": 1.8943,
       "step": 2158
     },
     {
       "epoch": 3.4544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9117224216461182,
+      "learning_rate": 0.0005788499999999999,
+      "loss": 2.1518,
       "step": 2159
     },
     {
       "epoch": 3.456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7245050072669983,
+      "learning_rate": 0.0005786999999999999,
+      "loss": 1.9638,
       "step": 2160
     },
     {
       "epoch": 3.4576000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.124849319458008,
+      "learning_rate": 0.0005785499999999999,
+      "loss": 1.8997,
       "step": 2161
     },
     {
       "epoch": 3.4592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2298882007598877,
+      "learning_rate": 0.0005784,
+      "loss": 2.3108,
       "step": 2162
     },
     {
       "epoch": 3.4608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4195408821105957,
+      "learning_rate": 0.00057825,
+      "loss": 1.9164,
       "step": 2163
     },
     {
       "epoch": 3.4624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2068179845809937,
+      "learning_rate": 0.0005781,
+      "loss": 2.1875,
       "step": 2164
     },
     {
       "epoch": 3.464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8306836485862732,
+      "learning_rate": 0.0005779499999999999,
+      "loss": 2.1143,
       "step": 2165
     },
     {
       "epoch": 3.4656000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5825016498565674,
+      "learning_rate": 0.0005778,
+      "loss": 2.0437,
       "step": 2166
     },
     {
       "epoch": 3.4672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7154268622398376,
+      "learning_rate": 0.00057765,
+      "loss": 1.8824,
       "step": 2167
     },
     {
       "epoch": 3.4688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1212105751037598,
+      "learning_rate": 0.0005775,
+      "loss": 2.2052,
       "step": 2168
     },
     {
       "epoch": 3.4704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1738407611846924,
+      "learning_rate": 0.00057735,
+      "loss": 2.0513,
       "step": 2169
     },
     {
       "epoch": 3.472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4919712543487549,
+      "learning_rate": 0.0005771999999999999,
+      "loss": 1.8867,
       "step": 2170
     },
     {
       "epoch": 3.4736000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7555265426635742,
+      "learning_rate": 0.00057705,
+      "loss": 1.8786,
       "step": 2171
     },
     {
       "epoch": 3.4752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0224101543426514,
+      "learning_rate": 0.0005769,
+      "loss": 2.0857,
       "step": 2172
     },
     {
       "epoch": 3.4768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5990490913391113,
+      "learning_rate": 0.00057675,
+      "loss": 2.1757,
       "step": 2173
     },
     {
       "epoch": 3.4784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5588961839675903,
+      "learning_rate": 0.0005765999999999999,
+      "loss": 2.0964,
       "step": 2174
     },
     {
       "epoch": 3.48,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.0005765999999999999,
+      "loss": 2.2852,
       "step": 2175
     },
     {
       "epoch": 3.4816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 18.246580123901367,
+      "learning_rate": 0.0005764499999999999,
+      "loss": 3.2206,
       "step": 2176
     },
     {
       "epoch": 3.4832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.455694198608398,
+      "learning_rate": 0.0005763,
+      "loss": 2.588,
       "step": 2177
     },
     {
       "epoch": 3.4848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2639204263687134,
+      "learning_rate": 0.00057615,
+      "loss": 2.1951,
       "step": 2178
     },
     {
       "epoch": 3.4864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.898868441581726,
+      "learning_rate": 0.0005759999999999999,
+      "loss": 2.1804,
       "step": 2179
     },
     {
       "epoch": 3.488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.800933301448822,
+      "learning_rate": 0.0005758499999999999,
+      "loss": 2.2862,
       "step": 2180
     },
     {
       "epoch": 3.4896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.313568592071533,
+      "learning_rate": 0.0005757,
+      "loss": 2.2126,
       "step": 2181
     },
     {
       "epoch": 3.4912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0708627700805664,
+      "learning_rate": 0.00057555,
+      "loss": 2.1175,
       "step": 2182
     },
     {
       "epoch": 3.4928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.972640037536621,
+      "learning_rate": 0.0005753999999999999,
+      "loss": 2.0464,
       "step": 2183
     },
     {
       "epoch": 3.4944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3649954795837402,
+      "learning_rate": 0.0005752499999999999,
+      "loss": 2.0686,
       "step": 2184
     },
     {
       "epoch": 3.496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0719057321548462,
+      "learning_rate": 0.0005750999999999999,
+      "loss": 2.0874,
       "step": 2185
     },
     {
       "epoch": 3.4976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8816778659820557,
+      "learning_rate": 0.00057495,
+      "loss": 2.1214,
       "step": 2186
     },
     {
       "epoch": 3.4992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8715779781341553,
+      "learning_rate": 0.0005747999999999999,
+      "loss": 1.887,
       "step": 2187
     },
     {
       "epoch": 3.5008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8262019753456116,
+      "learning_rate": 0.0005746499999999999,
+      "loss": 1.938,
       "step": 2188
     },
     {
       "epoch": 3.5023999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9271842241287231,
+      "learning_rate": 0.0005744999999999999,
+      "loss": 1.7774,
       "step": 2189
     },
     {
       "epoch": 3.504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.82608962059021,
+      "learning_rate": 0.00057435,
+      "loss": 1.8269,
       "step": 2190
     },
     {
       "epoch": 3.5056000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2712914943695068,
+      "learning_rate": 0.0005742,
+      "loss": 1.9094,
       "step": 2191
     },
     {
       "epoch": 3.5072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9681758880615234,
+      "learning_rate": 0.00057405,
+      "loss": 1.8966,
       "step": 2192
     },
     {
       "epoch": 3.5088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9789705276489258,
+      "learning_rate": 0.0005738999999999999,
+      "loss": 2.0215,
       "step": 2193
     },
     {
       "epoch": 3.5103999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8557264804840088,
+      "learning_rate": 0.0005737499999999999,
+      "loss": 2.0703,
       "step": 2194
     },
     {
       "epoch": 3.512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9306953549385071,
+      "learning_rate": 0.0005736,
+      "loss": 1.687,
       "step": 2195
     },
     {
       "epoch": 3.5136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0400307178497314,
+      "learning_rate": 0.00057345,
+      "loss": 1.9039,
       "step": 2196
     },
     {
       "epoch": 3.5152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0845005512237549,
+      "learning_rate": 0.0005733,
+      "loss": 1.8057,
       "step": 2197
     },
     {
       "epoch": 3.5168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5625959634780884,
+      "learning_rate": 0.00057315,
+      "loss": 1.7292,
       "step": 2198
     },
     {
       "epoch": 3.5183999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7412774562835693,
+      "learning_rate": 0.0005729999999999999,
+      "loss": 1.858,
       "step": 2199
     },
     {
       "epoch": 3.52,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2200
-    },
-    {
-      "epoch": 3.52,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 157.7226,
-      "eval_samples_per_second": 19.883,
-      "eval_steps_per_second": 1.243,
-      "eval_wer": 1.0,
+      "grad_norm": 0.7688709497451782,
+      "learning_rate": 0.00057285,
+      "loss": 1.9147,
       "step": 2200
     },
     {
       "epoch": 3.5216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7633388638496399,
+      "learning_rate": 0.0005727,
+      "loss": 1.9017,
       "step": 2201
     },
     {
       "epoch": 3.5232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4895182251930237,
+      "learning_rate": 0.00057255,
+      "loss": 1.8924,
       "step": 2202
     },
     {
       "epoch": 3.5248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5399225950241089,
+      "learning_rate": 0.0005723999999999999,
+      "loss": 1.6966,
       "step": 2203
     },
     {
       "epoch": 3.5263999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.073164701461792,
+      "learning_rate": 0.00057225,
+      "loss": 1.7686,
       "step": 2204
     },
     {
       "epoch": 3.528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9257833957672119,
+      "learning_rate": 0.0005721,
+      "loss": 1.9916,
       "step": 2205
     },
     {
       "epoch": 3.5296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.169154405593872,
+      "learning_rate": 0.00057195,
+      "loss": 2.0111,
       "step": 2206
     },
     {
       "epoch": 3.5312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2770828008651733,
+      "learning_rate": 0.0005717999999999999,
+      "loss": 1.6243,
       "step": 2207
     },
     {
       "epoch": 3.5328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6825202703475952,
+      "learning_rate": 0.0005716499999999999,
+      "loss": 1.8726,
       "step": 2208
     },
     {
       "epoch": 3.5343999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7958777546882629,
+      "learning_rate": 0.0005715,
+      "loss": 1.7756,
       "step": 2209
     },
     {
       "epoch": 3.536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5408828258514404,
+      "learning_rate": 0.00057135,
+      "loss": 1.9524,
       "step": 2210
     },
     {
       "epoch": 3.5376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8456842303276062,
+      "learning_rate": 0.0005711999999999999,
+      "loss": 2.0548,
       "step": 2211
     },
     {
       "epoch": 3.5392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.612095057964325,
+      "learning_rate": 0.0005710499999999999,
+      "loss": 2.2309,
       "step": 2212
     },
     {
       "epoch": 3.5408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2765395641326904,
+      "learning_rate": 0.0005708999999999999,
+      "loss": 2.3198,
       "step": 2213
     },
     {
       "epoch": 3.5423999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.418147325515747,
+      "learning_rate": 0.00057075,
+      "loss": 1.9075,
       "step": 2214
     },
     {
       "epoch": 3.544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.177223205566406,
+      "learning_rate": 0.0005705999999999999,
+      "loss": 2.0796,
       "step": 2215
     },
     {
       "epoch": 3.5456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.717703938484192,
+      "learning_rate": 0.0005704499999999999,
+      "loss": 2.3406,
       "step": 2216
     },
     {
       "epoch": 3.5472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.981576681137085,
+      "learning_rate": 0.0005702999999999999,
+      "loss": 2.2258,
       "step": 2217
     },
     {
       "epoch": 3.5488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4039616584777832,
+      "learning_rate": 0.00057015,
+      "loss": 1.7069,
       "step": 2218
     },
     {
       "epoch": 3.5504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3874764442443848,
+      "learning_rate": 0.00057,
+      "loss": 1.9812,
       "step": 2219
     },
     {
       "epoch": 3.552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5490448474884033,
+      "learning_rate": 0.0005698499999999999,
+      "loss": 1.9941,
       "step": 2220
     },
     {
       "epoch": 3.5536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8983708620071411,
+      "learning_rate": 0.0005696999999999999,
+      "loss": 2.2199,
       "step": 2221
     },
     {
       "epoch": 3.5552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5205702781677246,
+      "learning_rate": 0.0005695499999999999,
+      "loss": 2.207,
       "step": 2222
     },
     {
       "epoch": 3.5568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4460971355438232,
+      "learning_rate": 0.0005694,
+      "loss": 2.485,
       "step": 2223
     },
     {
       "epoch": 3.5584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5268499851226807,
+      "learning_rate": 0.00056925,
+      "loss": 2.3391,
       "step": 2224
     },
     {
       "epoch": 3.56,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8089544773101807,
+      "learning_rate": 0.0005691,
+      "loss": 2.5278,
       "step": 2225
     },
     {
       "epoch": 3.5616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 14.736753463745117,
+      "learning_rate": 0.0005689499999999999,
+      "loss": 2.7849,
       "step": 2226
     },
     {
       "epoch": 3.5632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.6178412437438965,
+      "learning_rate": 0.0005688,
+      "loss": 2.3745,
       "step": 2227
     },
     {
       "epoch": 3.5648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.448739051818848,
+      "learning_rate": 0.00056865,
+      "loss": 2.56,
       "step": 2228
     },
     {
       "epoch": 3.5664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.70924711227417,
+      "learning_rate": 0.0005685,
+      "loss": 2.4006,
       "step": 2229
     },
     {
       "epoch": 3.568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.747269153594971,
+      "learning_rate": 0.00056835,
+      "loss": 2.1758,
       "step": 2230
     },
     {
       "epoch": 3.5696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0769402980804443,
+      "learning_rate": 0.0005681999999999999,
+      "loss": 2.0656,
       "step": 2231
     },
     {
       "epoch": 3.5712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1517651081085205,
+      "learning_rate": 0.00056805,
+      "loss": 2.0493,
       "step": 2232
     },
     {
       "epoch": 3.5728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1098272800445557,
+      "learning_rate": 0.0005679,
+      "loss": 1.889,
       "step": 2233
     },
     {
       "epoch": 3.5744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1274043321609497,
+      "learning_rate": 0.00056775,
+      "loss": 1.9614,
       "step": 2234
     },
     {
       "epoch": 3.576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.16263484954834,
+      "learning_rate": 0.0005675999999999999,
+      "loss": 1.9668,
       "step": 2235
     },
     {
       "epoch": 3.5776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4705198407173157,
+      "learning_rate": 0.0005674499999999999,
+      "loss": 1.9278,
       "step": 2236
     },
     {
       "epoch": 3.5792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4408669471740723,
+      "learning_rate": 0.0005673,
+      "loss": 1.8853,
       "step": 2237
     },
     {
       "epoch": 3.5808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9206117987632751,
+      "learning_rate": 0.00056715,
+      "loss": 1.9511,
       "step": 2238
     },
     {
       "epoch": 3.5824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8084490299224854,
+      "learning_rate": 0.0005669999999999999,
+      "loss": 1.7235,
       "step": 2239
     },
     {
       "epoch": 3.584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9082566499710083,
+      "learning_rate": 0.0005668499999999999,
+      "loss": 1.6991,
       "step": 2240
     },
     {
       "epoch": 3.5856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1258739233016968,
+      "learning_rate": 0.0005667,
+      "loss": 1.8765,
       "step": 2241
     },
     {
       "epoch": 3.5872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7276796102523804,
+      "learning_rate": 0.00056655,
+      "loss": 1.6729,
       "step": 2242
     },
     {
       "epoch": 3.5888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.931858539581299,
+      "learning_rate": 0.0005663999999999999,
+      "loss": 1.746,
       "step": 2243
     },
     {
       "epoch": 3.5904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8679087162017822,
+      "learning_rate": 0.0005662499999999999,
+      "loss": 2.0839,
       "step": 2244
     },
     {
       "epoch": 3.592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.481023073196411,
+      "learning_rate": 0.0005660999999999999,
+      "loss": 1.8341,
       "step": 2245
     },
     {
       "epoch": 3.5936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.399448871612549,
+      "learning_rate": 0.00056595,
+      "loss": 1.9251,
       "step": 2246
     },
     {
       "epoch": 3.5952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5826406478881836,
+      "learning_rate": 0.0005657999999999999,
+      "loss": 1.6982,
       "step": 2247
     },
     {
       "epoch": 3.5968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7085354328155518,
+      "learning_rate": 0.0005656499999999999,
+      "loss": 1.8364,
       "step": 2248
     },
     {
       "epoch": 3.5984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.819995641708374,
+      "learning_rate": 0.0005654999999999999,
+      "loss": 2.0573,
       "step": 2249
     },
     {
       "epoch": 3.6,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9627421498298645,
+      "learning_rate": 0.0005653499999999999,
+      "loss": 1.8237,
       "step": 2250
     },
     {
       "epoch": 3.6016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1660212278366089,
+      "learning_rate": 0.0005652,
+      "loss": 2.1266,
       "step": 2251
     },
     {
       "epoch": 3.6032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0879263877868652,
+      "learning_rate": 0.0005650499999999999,
+      "loss": 1.9085,
       "step": 2252
     },
     {
       "epoch": 3.6048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3968796730041504,
+      "learning_rate": 0.0005648999999999999,
+      "loss": 1.7709,
       "step": 2253
     },
     {
       "epoch": 3.6064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.195981025695801,
+      "learning_rate": 0.0005647499999999999,
+      "loss": 1.8685,
       "step": 2254
     },
     {
       "epoch": 3.608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6781668663024902,
+      "learning_rate": 0.0005646,
+      "loss": 1.8469,
       "step": 2255
     },
     {
       "epoch": 3.6096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7052363157272339,
+      "learning_rate": 0.00056445,
+      "loss": 1.9456,
       "step": 2256
     },
     {
       "epoch": 3.6112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.590246558189392,
+      "learning_rate": 0.0005643,
+      "loss": 2.0146,
       "step": 2257
     },
     {
       "epoch": 3.6128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8339739441871643,
+      "learning_rate": 0.0005641499999999999,
+      "loss": 1.8619,
       "step": 2258
     },
     {
       "epoch": 3.6144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.960278272628784,
+      "learning_rate": 0.0005639999999999999,
+      "loss": 1.8805,
       "step": 2259
     },
     {
       "epoch": 3.616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9200167655944824,
+      "learning_rate": 0.00056385,
+      "loss": 1.7948,
       "step": 2260
     },
     {
       "epoch": 3.6176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0074909925460815,
+      "learning_rate": 0.0005637,
+      "loss": 1.8153,
       "step": 2261
     },
     {
       "epoch": 3.6192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8568616509437561,
+      "learning_rate": 0.00056355,
+      "loss": 2.0004,
       "step": 2262
     },
     {
       "epoch": 3.6208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.956713080406189,
+      "learning_rate": 0.0005633999999999999,
+      "loss": 2.0911,
       "step": 2263
     },
     {
       "epoch": 3.6224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9284929037094116,
+      "learning_rate": 0.00056325,
+      "loss": 1.8288,
       "step": 2264
     },
     {
       "epoch": 3.624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.894782543182373,
+      "learning_rate": 0.0005631,
+      "loss": 1.9543,
       "step": 2265
     },
     {
       "epoch": 3.6256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8398752212524414,
+      "learning_rate": 0.00056295,
+      "loss": 2.1485,
       "step": 2266
     },
     {
       "epoch": 3.6272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.454357147216797,
+      "learning_rate": 0.0005627999999999999,
+      "loss": 2.0567,
       "step": 2267
     },
     {
       "epoch": 3.6288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.641742467880249,
+      "learning_rate": 0.0005626499999999999,
+      "loss": 2.2341,
       "step": 2268
     },
     {
       "epoch": 3.6304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7430512309074402,
+      "learning_rate": 0.0005625,
+      "loss": 2.0746,
       "step": 2269
     },
     {
       "epoch": 3.632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.865148663520813,
+      "learning_rate": 0.00056235,
+      "loss": 2.1452,
       "step": 2270
     },
     {
       "epoch": 3.6336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5117383003234863,
+      "learning_rate": 0.0005622,
+      "loss": 1.9833,
       "step": 2271
     },
     {
       "epoch": 3.6352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9610617756843567,
+      "learning_rate": 0.0005620499999999999,
+      "loss": 2.2757,
       "step": 2272
     },
     {
       "epoch": 3.6368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3695871829986572,
+      "learning_rate": 0.0005618999999999999,
+      "loss": 2.0467,
       "step": 2273
     },
     {
       "epoch": 3.6384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8381495475769043,
+      "learning_rate": 0.00056175,
+      "loss": 2.2694,
       "step": 2274
     },
     {
       "epoch": 3.64,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5025103092193604,
+      "learning_rate": 0.0005616,
+      "loss": 2.7829,
       "step": 2275
     },
     {
       "epoch": 3.6416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.791909217834473,
+      "learning_rate": 0.0005614499999999999,
+      "loss": 2.5232,
       "step": 2276
     },
     {
       "epoch": 3.6432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.33229660987854,
+      "learning_rate": 0.0005612999999999999,
+      "loss": 1.9347,
       "step": 2277
     },
     {
       "epoch": 3.6448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.1871113777160645,
+      "learning_rate": 0.00056115,
+      "loss": 2.4298,
       "step": 2278
     },
     {
       "epoch": 3.6464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.01583194732666,
+      "learning_rate": 0.000561,
+      "loss": 2.3458,
       "step": 2279
     },
     {
       "epoch": 3.648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6337729692459106,
+      "learning_rate": 0.0005608499999999999,
+      "loss": 2.2367,
       "step": 2280
     },
     {
       "epoch": 3.6496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.28971004486084,
+      "learning_rate": 0.0005606999999999999,
+      "loss": 2.3382,
       "step": 2281
     },
     {
       "epoch": 3.6512000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2333521842956543,
+      "learning_rate": 0.0005605499999999999,
+      "loss": 2.0587,
       "step": 2282
     },
     {
       "epoch": 3.6528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1958487033843994,
+      "learning_rate": 0.0005604,
+      "loss": 1.8286,
       "step": 2283
     },
     {
       "epoch": 3.6544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8140027523040771,
+      "learning_rate": 0.00056025,
+      "loss": 1.9269,
       "step": 2284
     },
     {
       "epoch": 3.656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.204829216003418,
+      "learning_rate": 0.0005600999999999999,
+      "loss": 1.7265,
       "step": 2285
     },
     {
       "epoch": 3.6576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2901850938796997,
+      "learning_rate": 0.0005599499999999999,
+      "loss": 2.0729,
       "step": 2286
     },
     {
       "epoch": 3.6592000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9441965818405151,
+      "learning_rate": 0.0005598,
+      "loss": 1.7402,
       "step": 2287
     },
     {
       "epoch": 3.6608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9471627473831177,
+      "learning_rate": 0.00055965,
+      "loss": 1.9046,
       "step": 2288
     },
     {
       "epoch": 3.6624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4211928844451904,
+      "learning_rate": 0.0005595,
+      "loss": 1.7967,
       "step": 2289
     },
     {
       "epoch": 3.664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5628670454025269,
+      "learning_rate": 0.00055935,
+      "loss": 1.5733,
       "step": 2290
     },
     {
       "epoch": 3.6656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8597891330718994,
+      "learning_rate": 0.0005591999999999999,
+      "loss": 1.8503,
       "step": 2291
     },
     {
       "epoch": 3.6672000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4017963409423828,
+      "learning_rate": 0.00055905,
+      "loss": 1.8248,
       "step": 2292
     },
     {
       "epoch": 3.6688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8284274339675903,
+      "learning_rate": 0.0005589,
+      "loss": 1.9808,
       "step": 2293
     },
     {
       "epoch": 3.6704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6716489791870117,
+      "learning_rate": 0.00055875,
+      "loss": 1.756,
       "step": 2294
     },
     {
       "epoch": 3.672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6753265857696533,
+      "learning_rate": 0.0005586,
+      "loss": 1.9613,
       "step": 2295
     },
     {
       "epoch": 3.6736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1374127864837646,
+      "learning_rate": 0.0005584499999999999,
+      "loss": 1.699,
       "step": 2296
     },
     {
       "epoch": 3.6752000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8114046454429626,
+      "learning_rate": 0.0005583,
+      "loss": 1.8101,
       "step": 2297
     },
     {
       "epoch": 3.6768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2519382238388062,
+      "learning_rate": 0.00055815,
+      "loss": 1.9894,
       "step": 2298
     },
     {
       "epoch": 3.6784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3280014991760254,
+      "learning_rate": 0.000558,
+      "loss": 1.6685,
       "step": 2299
     },
     {
       "epoch": 3.68,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2300
-    },
-    {
-      "epoch": 3.68,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 158.3126,
-      "eval_samples_per_second": 19.809,
-      "eval_steps_per_second": 1.238,
-      "eval_wer": 1.0,
+      "grad_norm": 0.5929081439971924,
+      "learning_rate": 0.0005578499999999999,
+      "loss": 1.8905,
       "step": 2300
     },
     {
       "epoch": 3.6816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.48221904039382935,
+      "learning_rate": 0.0005577,
+      "loss": 1.9267,
       "step": 2301
     },
     {
       "epoch": 3.6832000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7531263828277588,
+      "learning_rate": 0.00055755,
+      "loss": 1.6909,
       "step": 2302
     },
     {
       "epoch": 3.6848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1919063329696655,
+      "learning_rate": 0.0005574,
+      "loss": 1.6607,
       "step": 2303
     },
     {
       "epoch": 3.6864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6339486837387085,
+      "learning_rate": 0.0005572499999999999,
+      "loss": 2.0789,
       "step": 2304
     },
     {
       "epoch": 3.6879999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7811375260353088,
+      "learning_rate": 0.0005570999999999999,
+      "loss": 2.0737,
       "step": 2305
     },
     {
       "epoch": 3.6896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5296683311462402,
+      "learning_rate": 0.00055695,
+      "loss": 2.0396,
       "step": 2306
     },
     {
       "epoch": 3.6912000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6752074360847473,
+      "learning_rate": 0.0005568,
+      "loss": 1.8817,
       "step": 2307
     },
     {
       "epoch": 3.6928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5748279094696045,
+      "learning_rate": 0.0005566499999999999,
+      "loss": 1.7463,
       "step": 2308
     },
     {
       "epoch": 3.6944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6146976947784424,
+      "learning_rate": 0.0005564999999999999,
+      "loss": 1.7846,
       "step": 2309
     },
     {
       "epoch": 3.6959999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5607794523239136,
+      "learning_rate": 0.0005563499999999999,
+      "loss": 2.0738,
       "step": 2310
     },
     {
       "epoch": 3.6976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7432671785354614,
+      "learning_rate": 0.0005562,
+      "loss": 1.8349,
       "step": 2311
     },
     {
       "epoch": 3.6992000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9773478507995605,
+      "learning_rate": 0.0005560499999999999,
+      "loss": 2.1008,
       "step": 2312
     },
     {
       "epoch": 3.7008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4571824073791504,
+      "learning_rate": 0.0005558999999999999,
+      "loss": 2.1606,
       "step": 2313
     },
     {
       "epoch": 3.7024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5633294582366943,
+      "learning_rate": 0.0005557499999999999,
+      "loss": 1.7593,
       "step": 2314
     },
     {
       "epoch": 3.7039999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5776718258857727,
+      "learning_rate": 0.0005556,
+      "loss": 1.8431,
       "step": 2315
     },
     {
       "epoch": 3.7056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6635186672210693,
+      "learning_rate": 0.00055545,
+      "loss": 2.0499,
       "step": 2316
     },
     {
       "epoch": 3.7072000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.93550968170166,
+      "learning_rate": 0.0005552999999999999,
+      "loss": 2.1281,
       "step": 2317
     },
     {
       "epoch": 3.7088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0490763187408447,
+      "learning_rate": 0.0005551499999999999,
+      "loss": 1.8437,
       "step": 2318
     },
     {
       "epoch": 3.7104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8951698541641235,
+      "learning_rate": 0.0005549999999999999,
+      "loss": 2.1411,
       "step": 2319
     },
     {
       "epoch": 3.7119999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.894776463508606,
+      "learning_rate": 0.00055485,
+      "loss": 2.3416,
       "step": 2320
     },
     {
       "epoch": 3.7136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8387718796730042,
+      "learning_rate": 0.0005547,
+      "loss": 2.3502,
       "step": 2321
     },
     {
       "epoch": 3.7152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7882157564163208,
+      "learning_rate": 0.00055455,
+      "loss": 2.3886,
       "step": 2322
     },
     {
       "epoch": 3.7168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4860434532165527,
+      "learning_rate": 0.0005544,
+      "loss": 1.9161,
       "step": 2323
     },
     {
       "epoch": 3.7184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8020563125610352,
+      "learning_rate": 0.00055425,
+      "loss": 1.8318,
       "step": 2324
     },
     {
       "epoch": 3.7199999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.03687047958374,
+      "learning_rate": 0.0005541,
+      "loss": 2.7315,
       "step": 2325
     },
     {
       "epoch": 3.7216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.88319993019104,
+      "learning_rate": 0.00055395,
+      "loss": 2.3845,
       "step": 2326
     },
     {
       "epoch": 3.7232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.421705722808838,
+      "learning_rate": 0.0005538,
+      "loss": 2.0725,
       "step": 2327
     },
     {
       "epoch": 3.7248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.533794403076172,
+      "learning_rate": 0.0005536499999999999,
+      "loss": 2.5097,
       "step": 2328
     },
     {
       "epoch": 3.7264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7621231079101562,
+      "learning_rate": 0.0005535,
+      "loss": 1.9029,
       "step": 2329
     },
     {
       "epoch": 3.7279999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.015029430389404,
+      "learning_rate": 0.00055335,
+      "loss": 2.3865,
       "step": 2330
     },
     {
       "epoch": 3.7296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4987126588821411,
+      "learning_rate": 0.0005532,
+      "loss": 2.0924,
       "step": 2331
     },
     {
       "epoch": 3.7312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7394864559173584,
+      "learning_rate": 0.0005530499999999999,
+      "loss": 1.9813,
       "step": 2332
     },
     {
       "epoch": 3.7328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7396926879882812,
+      "learning_rate": 0.0005528999999999999,
+      "loss": 2.0473,
       "step": 2333
     },
     {
       "epoch": 3.7344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7807384133338928,
+      "learning_rate": 0.00055275,
+      "loss": 1.8939,
       "step": 2334
     },
     {
       "epoch": 3.7359999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.359540939331055,
+      "learning_rate": 0.0005526,
+      "loss": 2.2214,
       "step": 2335
     },
     {
       "epoch": 3.7376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0581278800964355,
+      "learning_rate": 0.0005524499999999999,
+      "loss": 1.8789,
       "step": 2336
     },
     {
       "epoch": 3.7392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1891355514526367,
+      "learning_rate": 0.0005522999999999999,
+      "loss": 1.8932,
       "step": 2337
     },
     {
       "epoch": 3.7408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2215185165405273,
+      "learning_rate": 0.00055215,
+      "loss": 2.0718,
       "step": 2338
     },
     {
       "epoch": 3.7424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1517174243927,
+      "learning_rate": 0.000552,
+      "loss": 1.9519,
       "step": 2339
     },
     {
       "epoch": 3.7439999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9815576076507568,
+      "learning_rate": 0.0005518499999999999,
+      "loss": 2.1945,
       "step": 2340
     },
     {
       "epoch": 3.7456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4585610628128052,
+      "learning_rate": 0.0005516999999999999,
+      "loss": 1.7822,
       "step": 2341
     },
     {
       "epoch": 3.7472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6124539971351624,
+      "learning_rate": 0.0005515499999999999,
+      "loss": 1.8715,
       "step": 2342
     },
     {
       "epoch": 3.7488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5956341624259949,
+      "learning_rate": 0.0005514,
+      "loss": 1.7431,
       "step": 2343
     },
     {
       "epoch": 3.7504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5568941831588745,
+      "learning_rate": 0.0005512499999999999,
+      "loss": 1.6602,
       "step": 2344
     },
     {
       "epoch": 3.752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7756979465484619,
+      "learning_rate": 0.0005510999999999999,
+      "loss": 1.7601,
       "step": 2345
     },
     {
       "epoch": 3.7536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4702774286270142,
+      "learning_rate": 0.0005509499999999999,
+      "loss": 1.7106,
       "step": 2346
     },
     {
       "epoch": 3.7552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.420515239238739,
+      "learning_rate": 0.0005507999999999999,
+      "loss": 1.8738,
       "step": 2347
     },
     {
       "epoch": 3.7568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6160723567008972,
+      "learning_rate": 0.00055065,
+      "loss": 1.7276,
       "step": 2348
     },
     {
       "epoch": 3.7584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9224544167518616,
+      "learning_rate": 0.0005505,
+      "loss": 1.9053,
       "step": 2349
     },
     {
       "epoch": 3.76,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9273090362548828,
+      "learning_rate": 0.0005503499999999999,
+      "loss": 2.0036,
       "step": 2350
     },
     {
       "epoch": 3.7616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6562850475311279,
+      "learning_rate": 0.0005501999999999999,
+      "loss": 1.875,
       "step": 2351
     },
     {
       "epoch": 3.7632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0540215969085693,
+      "learning_rate": 0.00055005,
+      "loss": 1.6405,
       "step": 2352
     },
     {
       "epoch": 3.7648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9246771335601807,
+      "learning_rate": 0.0005499,
+      "loss": 1.6913,
       "step": 2353
     },
     {
       "epoch": 3.7664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6716334819793701,
+      "learning_rate": 0.00054975,
+      "loss": 1.7426,
       "step": 2354
     },
     {
       "epoch": 3.768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.676240861415863,
+      "learning_rate": 0.0005496,
+      "loss": 1.9785,
       "step": 2355
     },
     {
       "epoch": 3.7696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9691559076309204,
+      "learning_rate": 0.0005494499999999999,
+      "loss": 1.7538,
       "step": 2356
     },
     {
       "epoch": 3.7712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.520119309425354,
+      "learning_rate": 0.0005493,
+      "loss": 1.8912,
       "step": 2357
     },
     {
       "epoch": 3.7728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9442250728607178,
+      "learning_rate": 0.00054915,
+      "loss": 2.0052,
       "step": 2358
     },
     {
       "epoch": 3.7744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3105292320251465,
+      "learning_rate": 0.000549,
+      "loss": 2.0257,
       "step": 2359
     },
     {
       "epoch": 3.776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8196446895599365,
+      "learning_rate": 0.0005488499999999999,
+      "loss": 1.8295,
       "step": 2360
     },
     {
       "epoch": 3.7776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.038273334503174,
+      "learning_rate": 0.0005487,
+      "loss": 1.7075,
       "step": 2361
     },
     {
       "epoch": 3.7792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.246741771697998,
+      "learning_rate": 0.00054855,
+      "loss": 2.0361,
       "step": 2362
     },
     {
       "epoch": 3.7808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7947719693183899,
+      "learning_rate": 0.0005484,
+      "loss": 2.0181,
       "step": 2363
     },
     {
       "epoch": 3.7824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7120828032493591,
+      "learning_rate": 0.0005482499999999999,
+      "loss": 1.6842,
       "step": 2364
     },
     {
       "epoch": 3.784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5462130308151245,
+      "learning_rate": 0.0005480999999999999,
+      "loss": 1.9249,
       "step": 2365
     },
     {
       "epoch": 3.7856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.064439296722412,
+      "learning_rate": 0.00054795,
+      "loss": 2.1631,
       "step": 2366
     },
     {
       "epoch": 3.7872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4512231349945068,
+      "learning_rate": 0.0005478,
+      "loss": 1.8168,
       "step": 2367
     },
     {
       "epoch": 3.7888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4550942182540894,
+      "learning_rate": 0.0005476499999999999,
+      "loss": 1.8208,
       "step": 2368
     },
     {
       "epoch": 3.7904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.692131280899048,
+      "learning_rate": 0.0005474999999999999,
+      "loss": 2.1885,
       "step": 2369
     },
     {
       "epoch": 3.792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1041882038116455,
+      "learning_rate": 0.0005473499999999999,
+      "loss": 2.4662,
       "step": 2370
     },
     {
       "epoch": 3.7936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.566340684890747,
+      "learning_rate": 0.0005472,
+      "loss": 2.2271,
       "step": 2371
     },
     {
       "epoch": 3.7952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1705924272537231,
+      "learning_rate": 0.0005470499999999999,
+      "loss": 2.176,
       "step": 2372
     },
     {
       "epoch": 3.7968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.118368148803711,
+      "learning_rate": 0.0005468999999999999,
+      "loss": 1.7705,
       "step": 2373
     },
     {
       "epoch": 3.7984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.025905132293701,
+      "learning_rate": 0.0005467499999999999,
+      "loss": 2.2127,
       "step": 2374
     },
     {
       "epoch": 3.8,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.0005467499999999999,
+      "loss": 2.5841,
       "step": 2375
     },
     {
       "epoch": 3.8016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.623035430908203,
+      "learning_rate": 0.0005466,
+      "loss": 2.6382,
       "step": 2376
     },
     {
       "epoch": 3.8032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.104719161987305,
+      "learning_rate": 0.00054645,
+      "loss": 2.9251,
       "step": 2377
     },
     {
       "epoch": 3.8048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7435858845710754,
+      "learning_rate": 0.0005462999999999999,
+      "loss": 2.1993,
       "step": 2378
     },
     {
       "epoch": 3.8064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5204355716705322,
+      "learning_rate": 0.0005461499999999999,
+      "loss": 2.5489,
       "step": 2379
     },
     {
       "epoch": 3.808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5795775651931763,
+      "learning_rate": 0.0005459999999999999,
+      "loss": 2.2741,
       "step": 2380
     },
     {
       "epoch": 3.8096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9056549072265625,
+      "learning_rate": 0.00054585,
+      "loss": 1.968,
       "step": 2381
     },
     {
       "epoch": 3.8112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2619552612304688,
+      "learning_rate": 0.0005457,
+      "loss": 2.1325,
       "step": 2382
     },
     {
       "epoch": 3.8128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6153170466423035,
+      "learning_rate": 0.00054555,
+      "loss": 2.2253,
       "step": 2383
     },
     {
       "epoch": 3.8144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7306612730026245,
+      "learning_rate": 0.0005453999999999999,
+      "loss": 2.5022,
       "step": 2384
     },
     {
       "epoch": 3.816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6599177122116089,
+      "learning_rate": 0.0005452499999999999,
+      "loss": 1.8768,
       "step": 2385
     },
     {
       "epoch": 3.8176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.967004120349884,
+      "learning_rate": 0.0005451,
+      "loss": 2.0821,
       "step": 2386
     },
     {
       "epoch": 3.8192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.32717764377594,
+      "learning_rate": 0.00054495,
+      "loss": 1.6733,
       "step": 2387
     },
     {
       "epoch": 3.8208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4059408903121948,
+      "learning_rate": 0.0005448,
+      "loss": 1.9364,
       "step": 2388
     },
     {
       "epoch": 3.8224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1379108428955078,
+      "learning_rate": 0.0005446499999999999,
+      "loss": 1.742,
       "step": 2389
     },
     {
       "epoch": 3.824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.164423942565918,
+      "learning_rate": 0.0005445,
+      "loss": 1.9497,
       "step": 2390
     },
     {
       "epoch": 3.8256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.832740843296051,
+      "learning_rate": 0.00054435,
+      "loss": 1.7181,
       "step": 2391
     },
     {
       "epoch": 3.8272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2295310497283936,
+      "learning_rate": 0.0005442,
+      "loss": 1.8497,
       "step": 2392
     },
     {
       "epoch": 3.8288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0332822799682617,
+      "learning_rate": 0.0005440499999999999,
+      "loss": 1.8861,
       "step": 2393
     },
     {
       "epoch": 3.8304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0031321048736572,
+      "learning_rate": 0.0005438999999999999,
+      "loss": 1.7992,
       "step": 2394
     },
     {
       "epoch": 3.832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8428364396095276,
+      "learning_rate": 0.00054375,
+      "loss": 1.8351,
       "step": 2395
     },
     {
       "epoch": 3.8336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3388116359710693,
+      "learning_rate": 0.0005436,
+      "loss": 1.5866,
       "step": 2396
     },
     {
       "epoch": 3.8352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.911939859390259,
+      "learning_rate": 0.00054345,
+      "loss": 1.8575,
       "step": 2397
     },
     {
       "epoch": 3.8368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5401694178581238,
+      "learning_rate": 0.0005432999999999999,
+      "loss": 1.8289,
       "step": 2398
     },
     {
       "epoch": 3.8384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1643283367156982,
+      "learning_rate": 0.00054315,
+      "loss": 1.8512,
       "step": 2399
     },
     {
       "epoch": 3.84,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2400
-    },
-    {
-      "epoch": 3.84,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 158.5906,
-      "eval_samples_per_second": 19.774,
-      "eval_steps_per_second": 1.236,
-      "eval_wer": 1.0,
+      "grad_norm": 0.6926487684249878,
+      "learning_rate": 0.000543,
+      "loss": 1.762,
       "step": 2400
     },
     {
       "epoch": 3.8416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6354236006736755,
+      "learning_rate": 0.00054285,
+      "loss": 1.9255,
       "step": 2401
     },
     {
       "epoch": 3.8432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6662388443946838,
+      "learning_rate": 0.0005426999999999999,
+      "loss": 1.777,
       "step": 2402
     },
     {
       "epoch": 3.8448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7354152798652649,
+      "learning_rate": 0.0005425499999999999,
+      "loss": 1.7863,
       "step": 2403
     },
     {
       "epoch": 3.8464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2617355585098267,
+      "learning_rate": 0.0005424,
+      "loss": 1.6884,
       "step": 2404
     },
     {
       "epoch": 3.848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6013496518135071,
+      "learning_rate": 0.00054225,
+      "loss": 1.8549,
       "step": 2405
     },
     {
       "epoch": 3.8496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9593853950500488,
+      "learning_rate": 0.0005420999999999999,
+      "loss": 1.9073,
       "step": 2406
     },
     {
       "epoch": 3.8512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4911746978759766,
+      "learning_rate": 0.0005419499999999999,
+      "loss": 1.8498,
       "step": 2407
     },
     {
       "epoch": 3.8528000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5293194055557251,
+      "learning_rate": 0.0005417999999999999,
+      "loss": 2.0972,
       "step": 2408
     },
     {
       "epoch": 3.8544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1277806758880615,
+      "learning_rate": 0.00054165,
+      "loss": 1.5233,
       "step": 2409
     },
     {
       "epoch": 3.856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8829110264778137,
+      "learning_rate": 0.0005414999999999999,
+      "loss": 1.8209,
       "step": 2410
     },
     {
       "epoch": 3.8576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.754530668258667,
+      "learning_rate": 0.0005413499999999999,
+      "loss": 1.8229,
       "step": 2411
     },
     {
       "epoch": 3.8592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7509057521820068,
+      "learning_rate": 0.0005411999999999999,
+      "loss": 1.6897,
       "step": 2412
     },
     {
       "epoch": 3.8608000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7444249391555786,
+      "learning_rate": 0.00054105,
+      "loss": 1.761,
       "step": 2413
     },
     {
       "epoch": 3.8624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0965861082077026,
+      "learning_rate": 0.0005409,
+      "loss": 1.8965,
       "step": 2414
     },
     {
       "epoch": 3.864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9176476001739502,
+      "learning_rate": 0.00054075,
+      "loss": 1.7672,
       "step": 2415
     },
     {
       "epoch": 3.8656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6273837089538574,
+      "learning_rate": 0.0005405999999999999,
+      "loss": 2.1376,
       "step": 2416
     },
     {
       "epoch": 3.8672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.04819917678833,
+      "learning_rate": 0.0005404499999999999,
+      "loss": 1.6846,
       "step": 2417
     },
     {
       "epoch": 3.8688000000000002,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.0005404499999999999,
+      "loss": 0.9789,
       "step": 2418
     },
     {
       "epoch": 3.8704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5052590370178223,
+      "learning_rate": 0.0005403,
+      "loss": 2.0238,
       "step": 2419
     },
     {
       "epoch": 3.872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7614632248878479,
+      "learning_rate": 0.00054015,
+      "loss": 1.9592,
       "step": 2420
     },
     {
       "epoch": 3.8736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.096020460128784,
+      "learning_rate": 0.00054,
+      "loss": 2.0654,
       "step": 2421
     },
     {
       "epoch": 3.8752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6296525001525879,
+      "learning_rate": 0.00053985,
+      "loss": 1.896,
       "step": 2422
     },
     {
       "epoch": 3.8768000000000002,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4501912593841553,
+      "learning_rate": 0.0005396999999999999,
+      "loss": 2.1727,
       "step": 2423
     },
     {
       "epoch": 3.8784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7454720139503479,
+      "learning_rate": 0.00053955,
+      "loss": 2.0744,
       "step": 2424
     },
     {
       "epoch": 3.88,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1214377880096436,
+      "learning_rate": 0.0005394,
+      "loss": 2.232,
       "step": 2425
     },
     {
       "epoch": 3.8816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 25.60105323791504,
+      "learning_rate": 0.00053925,
+      "loss": 3.8323,
       "step": 2426
     },
     {
       "epoch": 3.8832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.382150888442993,
+      "learning_rate": 0.0005390999999999999,
+      "loss": 2.3454,
       "step": 2427
     },
     {
       "epoch": 3.8848000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5768654346466064,
+      "learning_rate": 0.00053895,
+      "loss": 2.6715,
       "step": 2428
     },
     {
       "epoch": 3.8864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9383482933044434,
+      "learning_rate": 0.0005388,
+      "loss": 2.5674,
       "step": 2429
     },
     {
       "epoch": 3.888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8392831087112427,
+      "learning_rate": 0.00053865,
+      "loss": 2.3529,
       "step": 2430
     },
     {
       "epoch": 3.8895999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.674063682556152,
+      "learning_rate": 0.0005384999999999999,
+      "loss": 2.0441,
       "step": 2431
     },
     {
       "epoch": 3.8912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.914694786071777,
+      "learning_rate": 0.0005383499999999999,
+      "loss": 2.2088,
       "step": 2432
     },
     {
       "epoch": 3.8928000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.900940418243408,
+      "learning_rate": 0.0005382,
+      "loss": 2.1332,
       "step": 2433
     },
     {
       "epoch": 3.8944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.249759674072266,
+      "learning_rate": 0.00053805,
+      "loss": 2.1121,
       "step": 2434
     },
     {
       "epoch": 3.896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5094621181488037,
+      "learning_rate": 0.0005378999999999999,
+      "loss": 1.928,
       "step": 2435
     },
     {
       "epoch": 3.8975999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.613595962524414,
+      "learning_rate": 0.0005377499999999999,
+      "loss": 2.1583,
       "step": 2436
     },
     {
       "epoch": 3.8992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.618670463562012,
+      "learning_rate": 0.0005376,
+      "loss": 2.1516,
       "step": 2437
     },
     {
       "epoch": 3.9008000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.50870943069458,
+      "learning_rate": 0.00053745,
+      "loss": 2.1445,
       "step": 2438
     },
     {
       "epoch": 3.9024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8594318628311157,
+      "learning_rate": 0.0005372999999999999,
+      "loss": 1.9006,
       "step": 2439
     },
     {
       "epoch": 3.904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5116372108459473,
+      "learning_rate": 0.0005371499999999999,
+      "loss": 1.8591,
       "step": 2440
     },
     {
       "epoch": 3.9055999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8958240747451782,
+      "learning_rate": 0.0005369999999999999,
+      "loss": 1.736,
       "step": 2441
     },
     {
       "epoch": 3.9072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5704692602157593,
+      "learning_rate": 0.00053685,
+      "loss": 1.7145,
       "step": 2442
     },
     {
       "epoch": 3.9088000000000003,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9452965259552002,
+      "learning_rate": 0.0005367,
+      "loss": 1.63,
       "step": 2443
     },
     {
       "epoch": 3.9104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.6801676750183105,
+      "learning_rate": 0.0005365499999999999,
+      "loss": 1.8781,
       "step": 2444
     },
     {
       "epoch": 3.912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4948861598968506,
+      "learning_rate": 0.0005363999999999999,
+      "loss": 1.8972,
       "step": 2445
     },
     {
       "epoch": 3.9135999999999997,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4566302299499512,
+      "learning_rate": 0.0005362499999999999,
+      "loss": 1.9978,
       "step": 2446
     },
     {
       "epoch": 3.9152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9597867131233215,
+      "learning_rate": 0.0005361,
+      "loss": 1.8151,
       "step": 2447
     },
     {
       "epoch": 3.9168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.585303544998169,
+      "learning_rate": 0.00053595,
+      "loss": 1.8052,
       "step": 2448
     },
     {
       "epoch": 3.9184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8317283391952515,
+      "learning_rate": 0.0005358,
+      "loss": 2.0335,
       "step": 2449
     },
     {
       "epoch": 3.92,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6935200691223145,
+      "learning_rate": 0.0005356499999999999,
+      "loss": 1.8919,
       "step": 2450
     },
     {
       "epoch": 3.9215999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9726048111915588,
+      "learning_rate": 0.0005355,
+      "loss": 1.7266,
       "step": 2451
     },
     {
       "epoch": 3.9232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2094123363494873,
+      "learning_rate": 0.00053535,
+      "loss": 1.815,
       "step": 2452
     },
     {
       "epoch": 3.9248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.217449426651001,
+      "learning_rate": 0.0005352,
+      "loss": 1.9153,
       "step": 2453
     },
     {
       "epoch": 3.9264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6052618622779846,
+      "learning_rate": 0.00053505,
+      "loss": 1.8094,
       "step": 2454
     },
     {
       "epoch": 3.928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7380341291427612,
+      "learning_rate": 0.0005348999999999999,
+      "loss": 1.7991,
       "step": 2455
     },
     {
       "epoch": 3.9295999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1621791124343872,
+      "learning_rate": 0.00053475,
+      "loss": 1.6631,
       "step": 2456
     },
     {
       "epoch": 3.9312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3532447814941406,
+      "learning_rate": 0.0005346,
+      "loss": 2.0495,
       "step": 2457
     },
     {
       "epoch": 3.9328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9457495808601379,
+      "learning_rate": 0.00053445,
+      "loss": 1.7621,
       "step": 2458
     },
     {
       "epoch": 3.9344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.45427045226097107,
+      "learning_rate": 0.0005342999999999999,
+      "loss": 1.7291,
       "step": 2459
     },
     {
       "epoch": 3.936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4032189846038818,
+      "learning_rate": 0.00053415,
+      "loss": 2.1577,
       "step": 2460
     },
     {
       "epoch": 3.9375999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6300065517425537,
+      "learning_rate": 0.000534,
+      "loss": 1.7902,
       "step": 2461
     },
     {
       "epoch": 3.9392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.3903911113739014,
+      "learning_rate": 0.00053385,
+      "loss": 1.8668,
       "step": 2462
     },
     {
       "epoch": 3.9408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4470239877700806,
+      "learning_rate": 0.0005336999999999999,
+      "loss": 1.9876,
       "step": 2463
     },
     {
       "epoch": 3.9424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.315370798110962,
+      "learning_rate": 0.0005335499999999999,
+      "loss": 1.9871,
       "step": 2464
     },
     {
       "epoch": 3.944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4395367503166199,
+      "learning_rate": 0.0005334,
+      "loss": 1.9859,
       "step": 2465
     },
     {
       "epoch": 3.9455999999999998,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0911147594451904,
+      "learning_rate": 0.00053325,
+      "loss": 1.6294,
       "step": 2466
     },
     {
       "epoch": 3.9472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0740833282470703,
+      "learning_rate": 0.0005330999999999999,
+      "loss": 1.7948,
       "step": 2467
     },
     {
       "epoch": 3.9488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.919568657875061,
+      "learning_rate": 0.0005329499999999999,
+      "loss": 1.8349,
       "step": 2468
     },
     {
       "epoch": 3.9504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3979318141937256,
+      "learning_rate": 0.0005327999999999999,
+      "loss": 1.6861,
       "step": 2469
     },
     {
       "epoch": 3.952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6491036415100098,
+      "learning_rate": 0.00053265,
+      "loss": 1.9822,
       "step": 2470
     },
     {
       "epoch": 3.9536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.751991868019104,
+      "learning_rate": 0.0005324999999999999,
+      "loss": 2.215,
       "step": 2471
     },
     {
       "epoch": 3.9552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5577988624572754,
+      "learning_rate": 0.0005323499999999999,
+      "loss": 2.007,
       "step": 2472
     },
     {
       "epoch": 3.9568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3308780193328857,
+      "learning_rate": 0.0005321999999999999,
+      "loss": 2.1599,
       "step": 2473
     },
     {
       "epoch": 3.9584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2450164556503296,
+      "learning_rate": 0.00053205,
+      "loss": 2.0319,
       "step": 2474
     },
     {
       "epoch": 3.96,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.967799186706543,
+      "learning_rate": 0.0005319,
+      "loss": 2.9772,
       "step": 2475
     },
     {
       "epoch": 3.9616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.091135263442993,
+      "learning_rate": 0.00053175,
+      "loss": 2.1827,
       "step": 2476
     },
     {
       "epoch": 3.9632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1590499877929688,
+      "learning_rate": 0.0005315999999999999,
+      "loss": 2.5292,
       "step": 2477
     },
     {
       "epoch": 3.9648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2901257276535034,
+      "learning_rate": 0.0005314499999999999,
+      "loss": 1.9916,
       "step": 2478
     },
     {
       "epoch": 3.9664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5774224400520325,
+      "learning_rate": 0.0005313,
+      "loss": 1.9182,
       "step": 2479
     },
     {
       "epoch": 3.968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7556675672531128,
+      "learning_rate": 0.00053115,
+      "loss": 1.7674,
       "step": 2480
     },
     {
       "epoch": 3.9696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9686545133590698,
+      "learning_rate": 0.000531,
+      "loss": 1.8127,
       "step": 2481
     },
     {
       "epoch": 3.9712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6765801906585693,
+      "learning_rate": 0.00053085,
+      "loss": 1.8753,
       "step": 2482
     },
     {
       "epoch": 3.9728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5046657919883728,
+      "learning_rate": 0.0005306999999999999,
+      "loss": 1.7438,
       "step": 2483
     },
     {
       "epoch": 3.9744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2890503406524658,
+      "learning_rate": 0.00053055,
+      "loss": 1.7384,
       "step": 2484
     },
     {
       "epoch": 3.976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7676361799240112,
+      "learning_rate": 0.0005304,
+      "loss": 1.7624,
       "step": 2485
     },
     {
       "epoch": 3.9776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7966771125793457,
+      "learning_rate": 0.00053025,
+      "loss": 1.6265,
       "step": 2486
     },
     {
       "epoch": 3.9792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9140196442604065,
+      "learning_rate": 0.0005300999999999999,
+      "loss": 1.8388,
       "step": 2487
     },
     {
       "epoch": 3.9808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5415698289871216,
+      "learning_rate": 0.00052995,
+      "loss": 2.0654,
       "step": 2488
     },
     {
       "epoch": 3.9824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7051250338554382,
+      "learning_rate": 0.0005298,
+      "loss": 1.6277,
       "step": 2489
     },
     {
       "epoch": 3.984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6715264916419983,
+      "learning_rate": 0.00052965,
+      "loss": 1.9315,
       "step": 2490
     },
     {
       "epoch": 3.9856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7733835577964783,
+      "learning_rate": 0.0005294999999999999,
+      "loss": 1.9635,
       "step": 2491
     },
     {
       "epoch": 3.9872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0471179485321045,
+      "learning_rate": 0.0005293499999999999,
+      "loss": 2.0734,
       "step": 2492
     },
     {
       "epoch": 3.9888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.750136137008667,
+      "learning_rate": 0.0005292,
+      "loss": 1.7293,
       "step": 2493
     },
     {
       "epoch": 3.9904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5307488441467285,
+      "learning_rate": 0.00052905,
+      "loss": 1.8903,
       "step": 2494
     },
     {
       "epoch": 3.992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4690338969230652,
+      "learning_rate": 0.0005288999999999999,
+      "loss": 1.8027,
       "step": 2495
     },
     {
       "epoch": 3.9936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9644952416419983,
+      "learning_rate": 0.0005287499999999999,
+      "loss": 1.7084,
       "step": 2496
     },
     {
       "epoch": 3.9952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2549068927764893,
+      "learning_rate": 0.0005286,
+      "loss": 1.8686,
       "step": 2497
     },
     {
       "epoch": 3.9968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7961026430130005,
+      "learning_rate": 0.00052845,
+      "loss": 1.9095,
       "step": 2498
     },
     {
       "epoch": 3.9984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.194058418273926,
+      "learning_rate": 0.0005282999999999999,
+      "loss": 2.1499,
       "step": 2499
     },
     {
       "epoch": 4.0,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2500
-    },
-    {
-      "epoch": 4.0,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 158.647,
-      "eval_samples_per_second": 19.767,
-      "eval_steps_per_second": 1.235,
-      "eval_wer": 1.0,
+      "learning_rate": 0.0005282999999999999,
+      "loss": 2.388,
       "step": 2500
     },
     {
       "epoch": 4.0016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.620855450630188,
+      "learning_rate": 0.0005281499999999999,
+      "loss": 2.3597,
       "step": 2501
     },
     {
       "epoch": 4.0032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7790733575820923,
+      "learning_rate": 0.0005279999999999999,
+      "loss": 2.3832,
       "step": 2502
     },
     {
       "epoch": 4.0048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5202178955078125,
+      "learning_rate": 0.00052785,
+      "loss": 1.8287,
       "step": 2503
     },
     {
       "epoch": 4.0064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0778002738952637,
+      "learning_rate": 0.0005276999999999999,
+      "loss": 2.0969,
       "step": 2504
     },
     {
       "epoch": 4.008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.827845335006714,
+      "learning_rate": 0.0005275499999999999,
+      "loss": 1.8303,
       "step": 2505
     },
     {
       "epoch": 4.0096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7472437620162964,
+      "learning_rate": 0.0005273999999999999,
+      "loss": 2.1248,
       "step": 2506
     },
     {
       "epoch": 4.0112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.076861381530762,
+      "learning_rate": 0.0005272499999999999,
+      "loss": 2.0531,
       "step": 2507
     },
     {
       "epoch": 4.0128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7650519013404846,
+      "learning_rate": 0.0005271,
+      "loss": 2.0667,
       "step": 2508
     },
     {
       "epoch": 4.0144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6342208385467529,
+      "learning_rate": 0.00052695,
+      "loss": 2.0258,
       "step": 2509
     },
     {
       "epoch": 4.016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6137911081314087,
+      "learning_rate": 0.0005267999999999999,
+      "loss": 2.0024,
       "step": 2510
     },
     {
       "epoch": 4.0176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.357777714729309,
+      "learning_rate": 0.0005266499999999999,
+      "loss": 1.7083,
       "step": 2511
     },
     {
       "epoch": 4.0192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2996630668640137,
+      "learning_rate": 0.0005265,
+      "loss": 1.8742,
       "step": 2512
     },
     {
       "epoch": 4.0208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.800222396850586,
+      "learning_rate": 0.00052635,
+      "loss": 1.8311,
       "step": 2513
     },
     {
       "epoch": 4.0224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6833199262619019,
+      "learning_rate": 0.0005262,
+      "loss": 1.8099,
       "step": 2514
     },
     {
       "epoch": 4.024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0427937507629395,
+      "learning_rate": 0.00052605,
+      "loss": 1.673,
       "step": 2515
     },
     {
       "epoch": 4.0256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4635120630264282,
+      "learning_rate": 0.0005258999999999999,
+      "loss": 1.7433,
       "step": 2516
     },
     {
       "epoch": 4.0272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7591537833213806,
+      "learning_rate": 0.00052575,
+      "loss": 1.9307,
       "step": 2517
     },
     {
       "epoch": 4.0288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.746607780456543,
+      "learning_rate": 0.0005256,
+      "loss": 2.116,
       "step": 2518
     },
     {
       "epoch": 4.0304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2699731588363647,
+      "learning_rate": 0.00052545,
+      "loss": 1.7969,
       "step": 2519
     },
     {
       "epoch": 4.032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5821665525436401,
+      "learning_rate": 0.0005252999999999999,
+      "loss": 1.8846,
       "step": 2520
     },
     {
       "epoch": 4.0336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0592167377471924,
+      "learning_rate": 0.0005251499999999999,
+      "loss": 1.7655,
       "step": 2521
     },
     {
       "epoch": 4.0352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6359199285507202,
+      "learning_rate": 0.000525,
+      "loss": 1.8698,
       "step": 2522
     },
     {
       "epoch": 4.0368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7579365968704224,
+      "learning_rate": 0.00052485,
+      "loss": 1.5655,
       "step": 2523
     },
     {
       "epoch": 4.0384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8781223297119141,
+      "learning_rate": 0.0005247,
+      "loss": 1.4955,
       "step": 2524
     },
     {
       "epoch": 4.04,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2994422912597656,
+      "learning_rate": 0.0005245499999999999,
+      "loss": 1.8925,
       "step": 2525
     },
     {
       "epoch": 4.0416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.251488447189331,
+      "learning_rate": 0.0005244,
+      "loss": 1.8904,
       "step": 2526
     },
     {
       "epoch": 4.0432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7512491941452026,
+      "learning_rate": 0.00052425,
+      "loss": 1.8389,
       "step": 2527
     },
     {
       "epoch": 4.0448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4011669158935547,
+      "learning_rate": 0.0005241,
+      "loss": 1.4594,
       "step": 2528
     },
     {
       "epoch": 4.0464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0186192989349365,
+      "learning_rate": 0.0005239499999999999,
+      "loss": 1.5813,
       "step": 2529
     },
     {
       "epoch": 4.048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7245293259620667,
+      "learning_rate": 0.0005237999999999999,
+      "loss": 1.8401,
       "step": 2530
     },
     {
       "epoch": 4.0496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.343597412109375,
+      "learning_rate": 0.00052365,
+      "loss": 2.2162,
       "step": 2531
     },
     {
       "epoch": 4.0512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9441453814506531,
+      "learning_rate": 0.0005235,
+      "loss": 1.87,
       "step": 2532
     },
     {
       "epoch": 4.0528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8596580624580383,
+      "learning_rate": 0.0005233499999999999,
+      "loss": 2.0299,
       "step": 2533
     },
     {
       "epoch": 4.0544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0425570011138916,
+      "learning_rate": 0.0005231999999999999,
+      "loss": 1.7958,
       "step": 2534
     },
     {
       "epoch": 4.056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5851573944091797,
+      "learning_rate": 0.00052305,
+      "loss": 2.0469,
       "step": 2535
     },
     {
       "epoch": 4.0576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8094300031661987,
+      "learning_rate": 0.0005229,
+      "loss": 1.9261,
       "step": 2536
     },
     {
       "epoch": 4.0592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7694682478904724,
+      "learning_rate": 0.0005227499999999999,
+      "loss": 1.8697,
       "step": 2537
     },
     {
       "epoch": 4.0608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.58815598487854,
+      "learning_rate": 0.0005225999999999999,
+      "loss": 2.1828,
       "step": 2538
     },
     {
       "epoch": 4.0624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3614821434020996,
+      "learning_rate": 0.0005224499999999999,
+      "loss": 1.8252,
       "step": 2539
     },
     {
       "epoch": 4.064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.085479497909546,
+      "learning_rate": 0.0005223,
+      "loss": 1.7422,
       "step": 2540
     },
     {
       "epoch": 4.0656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0933291912078857,
+      "learning_rate": 0.00052215,
+      "loss": 1.8915,
       "step": 2541
     },
     {
       "epoch": 4.0672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1529048681259155,
+      "learning_rate": 0.000522,
+      "loss": 1.9153,
       "step": 2542
     },
     {
       "epoch": 4.0688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6861165165901184,
+      "learning_rate": 0.0005218499999999999,
+      "loss": 2.0807,
       "step": 2543
     },
     {
       "epoch": 4.0704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9590578079223633,
+      "learning_rate": 0.0005216999999999999,
+      "loss": 1.6799,
       "step": 2544
     },
     {
       "epoch": 4.072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5041584968566895,
+      "learning_rate": 0.00052155,
+      "loss": 1.7953,
       "step": 2545
     },
     {
       "epoch": 4.0736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7498306035995483,
+      "learning_rate": 0.0005214,
+      "loss": 2.0022,
       "step": 2546
     },
     {
       "epoch": 4.0752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0625548362731934,
+      "learning_rate": 0.00052125,
+      "loss": 1.9091,
       "step": 2547
     },
     {
       "epoch": 4.0768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7577870488166809,
+      "learning_rate": 0.0005211,
+      "loss": 1.9865,
       "step": 2548
     },
     {
       "epoch": 4.0784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.3451220989227295,
+      "learning_rate": 0.00052095,
+      "loss": 2.2004,
       "step": 2549
     },
     {
       "epoch": 4.08,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8755617141723633,
+      "learning_rate": 0.0005208,
+      "loss": 2.4375,
       "step": 2550
     },
     {
       "epoch": 4.0816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 15.953272819519043,
+      "learning_rate": 0.00052065,
+      "loss": 3.0587,
       "step": 2551
     },
     {
       "epoch": 4.0832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.910642147064209,
+      "learning_rate": 0.0005205,
+      "loss": 2.5175,
       "step": 2552
     },
     {
       "epoch": 4.0848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.549471378326416,
+      "learning_rate": 0.0005203499999999999,
+      "loss": 2.3515,
       "step": 2553
     },
     {
       "epoch": 4.0864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.504085063934326,
+      "learning_rate": 0.0005202,
+      "loss": 2.2878,
       "step": 2554
     },
     {
       "epoch": 4.088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7265090942382812,
+      "learning_rate": 0.00052005,
+      "loss": 2.1208,
       "step": 2555
     },
     {
       "epoch": 4.0896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5239818096160889,
+      "learning_rate": 0.0005199,
+      "loss": 1.9764,
       "step": 2556
     },
     {
       "epoch": 4.0912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3443516492843628,
+      "learning_rate": 0.0005197499999999999,
+      "loss": 1.8913,
       "step": 2557
     },
     {
       "epoch": 4.0928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3654530048370361,
+      "learning_rate": 0.0005195999999999999,
+      "loss": 1.8563,
       "step": 2558
     },
     {
       "epoch": 4.0944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.616000652313232,
+      "learning_rate": 0.00051945,
+      "loss": 2.4474,
       "step": 2559
     },
     {
       "epoch": 4.096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2244317531585693,
+      "learning_rate": 0.0005193,
+      "loss": 2.0492,
       "step": 2560
     },
     {
       "epoch": 4.0976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2767560482025146,
+      "learning_rate": 0.0005191499999999999,
+      "loss": 1.9614,
       "step": 2561
     },
     {
       "epoch": 4.0992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6678037047386169,
+      "learning_rate": 0.0005189999999999999,
+      "loss": 1.9511,
       "step": 2562
     },
     {
       "epoch": 4.1008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.193560242652893,
+      "learning_rate": 0.00051885,
+      "loss": 1.6949,
       "step": 2563
     },
     {
       "epoch": 4.1024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6942158937454224,
+      "learning_rate": 0.0005187,
+      "loss": 1.6269,
       "step": 2564
     },
     {
       "epoch": 4.104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2404253482818604,
+      "learning_rate": 0.0005185499999999999,
+      "loss": 1.7762,
       "step": 2565
     },
     {
       "epoch": 4.1056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.07823646068573,
+      "learning_rate": 0.0005183999999999999,
+      "loss": 1.7197,
       "step": 2566
     },
     {
       "epoch": 4.1072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7477423548698425,
+      "learning_rate": 0.0005182499999999999,
+      "loss": 1.5912,
       "step": 2567
     },
     {
       "epoch": 4.1088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1995832920074463,
+      "learning_rate": 0.0005181,
+      "loss": 1.8297,
       "step": 2568
     },
     {
       "epoch": 4.1104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0321600437164307,
+      "learning_rate": 0.0005179499999999999,
+      "loss": 1.7202,
       "step": 2569
     },
     {
       "epoch": 4.112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0413057804107666,
+      "learning_rate": 0.0005177999999999999,
+      "loss": 1.6415,
       "step": 2570
     },
     {
       "epoch": 4.1136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.817716598510742,
+      "learning_rate": 0.0005176499999999999,
+      "loss": 1.7745,
       "step": 2571
     },
     {
       "epoch": 4.1152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.430630683898926,
+      "learning_rate": 0.0005175,
+      "loss": 1.96,
       "step": 2572
     },
     {
       "epoch": 4.1168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.758425712585449,
+      "learning_rate": 0.00051735,
+      "loss": 1.6291,
       "step": 2573
     },
     {
       "epoch": 4.1184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5142273902893066,
+      "learning_rate": 0.0005172,
+      "loss": 2.0679,
       "step": 2574
     },
     {
       "epoch": 4.12,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.235620141029358,
+      "learning_rate": 0.00051705,
+      "loss": 1.828,
       "step": 2575
     },
     {
       "epoch": 4.1216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9212775230407715,
+      "learning_rate": 0.0005168999999999999,
+      "loss": 1.7722,
       "step": 2576
     },
     {
       "epoch": 4.1232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6647447943687439,
+      "learning_rate": 0.00051675,
+      "loss": 1.906,
       "step": 2577
     },
     {
       "epoch": 4.1248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5452325344085693,
+      "learning_rate": 0.0005166,
+      "loss": 2.1333,
       "step": 2578
     },
     {
       "epoch": 4.1264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7366634011268616,
+      "learning_rate": 0.00051645,
+      "loss": 1.8803,
       "step": 2579
     },
     {
       "epoch": 4.128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7835949659347534,
+      "learning_rate": 0.0005163,
+      "loss": 1.9943,
       "step": 2580
     },
     {
       "epoch": 4.1296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.6397596001625061,
+      "learning_rate": 0.0005161499999999999,
+      "loss": 1.7529,
       "step": 2581
     },
     {
       "epoch": 4.1312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.079551935195923,
+      "learning_rate": 0.000516,
+      "loss": 1.9347,
       "step": 2582
     },
     {
       "epoch": 4.1328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6539299488067627,
+      "learning_rate": 0.00051585,
+      "loss": 1.7913,
       "step": 2583
     },
     {
       "epoch": 4.1344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2330212593078613,
+      "learning_rate": 0.0005157,
+      "loss": 2.1188,
       "step": 2584
     },
     {
       "epoch": 4.136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.305746078491211,
+      "learning_rate": 0.0005155499999999999,
+      "loss": 1.9467,
       "step": 2585
     },
     {
       "epoch": 4.1376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9980270862579346,
+      "learning_rate": 0.0005154,
+      "loss": 2.1646,
       "step": 2586
     },
     {
       "epoch": 4.1392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.092460036277771,
+      "learning_rate": 0.00051525,
+      "loss": 1.8353,
       "step": 2587
     },
     {
       "epoch": 4.1408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8863208889961243,
+      "learning_rate": 0.0005151,
+      "loss": 2.0309,
       "step": 2588
     },
     {
       "epoch": 4.1424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.000679850578308,
+      "learning_rate": 0.0005149499999999999,
+      "loss": 1.8691,
       "step": 2589
     },
     {
       "epoch": 4.144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8714463710784912,
+      "learning_rate": 0.0005147999999999999,
+      "loss": 1.8734,
       "step": 2590
     },
     {
       "epoch": 4.1456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2173759937286377,
+      "learning_rate": 0.00051465,
+      "loss": 2.3323,
       "step": 2591
     },
     {
       "epoch": 4.1472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8660641312599182,
+      "learning_rate": 0.0005145,
+      "loss": 2.1977,
       "step": 2592
     },
     {
       "epoch": 4.1488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4730968475341797,
+      "learning_rate": 0.0005143499999999999,
+      "loss": 2.1989,
       "step": 2593
     },
     {
       "epoch": 4.1504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7288036942481995,
+      "learning_rate": 0.0005141999999999999,
+      "loss": 1.7003,
       "step": 2594
     },
     {
       "epoch": 4.152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9436076283454895,
+      "learning_rate": 0.0005140499999999999,
+      "loss": 2.1883,
       "step": 2595
     },
     {
       "epoch": 4.1536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6240314245224,
+      "learning_rate": 0.0005139,
+      "loss": 1.9301,
       "step": 2596
     },
     {
       "epoch": 4.1552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.359818935394287,
+      "learning_rate": 0.0005137499999999999,
+      "loss": 2.1822,
       "step": 2597
     },
     {
       "epoch": 4.1568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.464782238006592,
+      "learning_rate": 0.0005135999999999999,
+      "loss": 2.2363,
       "step": 2598
     },
     {
       "epoch": 4.1584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7934373617172241,
+      "learning_rate": 0.0005134499999999999,
+      "loss": 2.5678,
       "step": 2599
     },
     {
       "epoch": 4.16,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2600
-    },
-    {
-      "epoch": 4.16,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 159.8763,
-      "eval_samples_per_second": 19.615,
-      "eval_steps_per_second": 1.226,
-      "eval_wer": 1.0,
+      "learning_rate": 0.0005134499999999999,
+      "loss": 2.6063,
       "step": 2600
     },
     {
       "epoch": 4.1616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.552441358566284,
+      "learning_rate": 0.0005133,
+      "loss": 1.8414,
       "step": 2601
     },
     {
       "epoch": 4.1632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0714163780212402,
+      "learning_rate": 0.00051315,
+      "loss": 2.1064,
       "step": 2602
     },
     {
       "epoch": 4.1648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.196826934814453,
+      "learning_rate": 0.0005129999999999999,
+      "loss": 2.618,
       "step": 2603
     },
     {
       "epoch": 4.1664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.336667060852051,
+      "learning_rate": 0.0005128499999999999,
+      "loss": 2.388,
       "step": 2604
     },
     {
       "epoch": 4.168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.398655891418457,
+      "learning_rate": 0.0005126999999999999,
+      "loss": 2.3414,
       "step": 2605
     },
     {
       "epoch": 4.1696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.885204315185547,
+      "learning_rate": 0.00051255,
+      "loss": 2.0417,
       "step": 2606
     },
     {
       "epoch": 4.1712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9722228050231934,
+      "learning_rate": 0.0005124,
+      "loss": 2.0161,
       "step": 2607
     },
     {
       "epoch": 4.1728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.073782444000244,
+      "learning_rate": 0.00051225,
+      "loss": 1.8719,
       "step": 2608
     },
     {
       "epoch": 4.1744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.423966884613037,
+      "learning_rate": 0.0005120999999999999,
+      "loss": 1.9612,
       "step": 2609
     },
     {
       "epoch": 4.176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.711324214935303,
+      "learning_rate": 0.00051195,
+      "loss": 2.8242,
       "step": 2610
     },
     {
       "epoch": 4.1776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0616002082824707,
+      "learning_rate": 0.0005118,
+      "loss": 1.7886,
       "step": 2611
     },
     {
       "epoch": 4.1792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0508960485458374,
+      "learning_rate": 0.00051165,
+      "loss": 1.651,
       "step": 2612
     },
     {
       "epoch": 4.1808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.420872211456299,
+      "learning_rate": 0.0005115,
+      "loss": 1.6546,
       "step": 2613
     },
     {
       "epoch": 4.1824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8311823606491089,
+      "learning_rate": 0.0005113499999999999,
+      "loss": 1.7491,
       "step": 2614
     },
     {
       "epoch": 4.184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1865077018737793,
+      "learning_rate": 0.0005112,
+      "loss": 1.8246,
       "step": 2615
     },
     {
       "epoch": 4.1856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.867258071899414,
+      "learning_rate": 0.00051105,
+      "loss": 1.8609,
       "step": 2616
     },
     {
       "epoch": 4.1872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9281314611434937,
+      "learning_rate": 0.0005109,
+      "loss": 1.7332,
       "step": 2617
     },
     {
       "epoch": 4.1888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.820015907287598,
+      "learning_rate": 0.0005107499999999999,
+      "loss": 1.9586,
       "step": 2618
     },
     {
       "epoch": 4.1904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.031200885772705,
+      "learning_rate": 0.0005105999999999999,
+      "loss": 1.5983,
       "step": 2619
     },
     {
       "epoch": 4.192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.678032398223877,
+      "learning_rate": 0.00051045,
+      "loss": 1.7672,
       "step": 2620
     },
     {
       "epoch": 4.1936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.8487162590026855,
+      "learning_rate": 0.0005103,
+      "loss": 2.5603,
       "step": 2621
     },
     {
       "epoch": 4.1952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8613600730895996,
+      "learning_rate": 0.0005101499999999999,
+      "loss": 1.747,
       "step": 2622
     },
     {
       "epoch": 4.1968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7120552659034729,
+      "learning_rate": 0.0005099999999999999,
+      "loss": 1.8262,
       "step": 2623
     },
     {
       "epoch": 4.1984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.790601968765259,
+      "learning_rate": 0.00050985,
+      "loss": 2.1263,
       "step": 2624
     },
     {
       "epoch": 4.2,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.589802622795105,
+      "learning_rate": 0.0005097,
+      "loss": 1.7034,
       "step": 2625
     },
     {
       "epoch": 4.2016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7992726564407349,
+      "learning_rate": 0.0005095499999999999,
+      "loss": 1.9422,
       "step": 2626
     },
     {
       "epoch": 4.2032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7839942574501038,
+      "learning_rate": 0.0005093999999999999,
+      "loss": 1.6971,
       "step": 2627
     },
     {
       "epoch": 4.2048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5990935564041138,
+      "learning_rate": 0.0005092499999999999,
+      "loss": 1.7538,
       "step": 2628
     },
     {
       "epoch": 4.2064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8034536838531494,
+      "learning_rate": 0.0005091,
+      "loss": 1.7504,
       "step": 2629
     },
     {
       "epoch": 4.208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9381115436553955,
+      "learning_rate": 0.0005089499999999999,
+      "loss": 2.0219,
       "step": 2630
     },
     {
       "epoch": 4.2096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.685932159423828,
+      "learning_rate": 0.0005087999999999999,
+      "loss": 1.7687,
       "step": 2631
     },
     {
       "epoch": 4.2112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8656719923019409,
+      "learning_rate": 0.0005086499999999999,
+      "loss": 1.5446,
       "step": 2632
     },
     {
       "epoch": 4.2128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.400219202041626,
+      "learning_rate": 0.0005085,
+      "loss": 2.0702,
       "step": 2633
     },
     {
       "epoch": 4.2144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.720746397972107,
+      "learning_rate": 0.00050835,
+      "loss": 2.0072,
       "step": 2634
     },
     {
       "epoch": 4.216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3523253202438354,
+      "learning_rate": 0.0005082,
+      "loss": 1.8526,
       "step": 2635
     },
     {
       "epoch": 4.2176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1961480379104614,
+      "learning_rate": 0.0005080499999999999,
+      "loss": 1.9631,
       "step": 2636
     },
     {
       "epoch": 4.2192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8764796257019043,
+      "learning_rate": 0.0005078999999999999,
+      "loss": 1.8485,
       "step": 2637
     },
     {
       "epoch": 4.2208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8735462427139282,
+      "learning_rate": 0.00050775,
+      "loss": 2.1159,
       "step": 2638
     },
     {
       "epoch": 4.2224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1814446449279785,
+      "learning_rate": 0.0005076,
+      "loss": 2.0074,
       "step": 2639
     },
     {
       "epoch": 4.224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8434025645256042,
+      "learning_rate": 0.00050745,
+      "loss": 1.7273,
       "step": 2640
     },
     {
       "epoch": 4.2256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.152543544769287,
+      "learning_rate": 0.0005073,
+      "loss": 2.4096,
       "step": 2641
     },
     {
       "epoch": 4.2272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9265046715736389,
+      "learning_rate": 0.0005071499999999999,
+      "loss": 2.0438,
       "step": 2642
     },
     {
       "epoch": 4.2288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5280780792236328,
+      "learning_rate": 0.000507,
+      "loss": 1.8898,
       "step": 2643
     },
     {
       "epoch": 4.2304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.905768096446991,
+      "learning_rate": 0.00050685,
+      "loss": 2.3321,
       "step": 2644
     },
     {
       "epoch": 4.232,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00050685,
+      "loss": 1.1,
       "step": 2645
     },
     {
       "epoch": 4.2336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8761323690414429,
+      "learning_rate": 0.0005067,
+      "loss": 1.9053,
       "step": 2646
     },
     {
       "epoch": 4.2352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.441887617111206,
+      "learning_rate": 0.0005065499999999999,
+      "loss": 2.1834,
       "step": 2647
     },
     {
       "epoch": 4.2368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4189505577087402,
+      "learning_rate": 0.0005064,
+      "loss": 2.2359,
       "step": 2648
     },
     {
       "epoch": 4.2384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.740108013153076,
+      "learning_rate": 0.00050625,
+      "loss": 2.0901,
       "step": 2649
     },
     {
       "epoch": 4.24,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9102271795272827,
+      "learning_rate": 0.0005061,
+      "loss": 2.6585,
       "step": 2650
     },
     {
       "epoch": 4.2416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 14.789002418518066,
+      "learning_rate": 0.00050595,
+      "loss": 2.8544,
       "step": 2651
     },
     {
       "epoch": 4.2432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1927456855773926,
+      "learning_rate": 0.0005057999999999999,
+      "loss": 2.0763,
       "step": 2652
     },
     {
       "epoch": 4.2448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.375650882720947,
+      "learning_rate": 0.00050565,
+      "loss": 2.3458,
       "step": 2653
     },
     {
       "epoch": 4.2464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2653048038482666,
+      "learning_rate": 0.0005055,
+      "loss": 2.3287,
       "step": 2654
     },
     {
       "epoch": 4.248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7491447925567627,
+      "learning_rate": 0.00050535,
+      "loss": 2.1713,
       "step": 2655
     },
     {
       "epoch": 4.2496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.120514392852783,
+      "learning_rate": 0.0005051999999999999,
+      "loss": 1.9272,
       "step": 2656
     },
     {
       "epoch": 4.2512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3644896745681763,
+      "learning_rate": 0.0005050499999999999,
+      "loss": 2.2318,
       "step": 2657
     },
     {
       "epoch": 4.2528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0127274990081787,
+      "learning_rate": 0.0005049,
+      "loss": 2.2203,
       "step": 2658
     },
     {
       "epoch": 4.2544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9749441146850586,
+      "learning_rate": 0.00050475,
+      "loss": 1.941,
       "step": 2659
     },
     {
       "epoch": 4.256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2064893245697021,
+      "learning_rate": 0.0005045999999999999,
+      "loss": 1.9854,
       "step": 2660
     },
     {
       "epoch": 4.2576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9935476779937744,
+      "learning_rate": 0.0005044499999999999,
+      "loss": 1.8353,
       "step": 2661
     },
     {
       "epoch": 4.2592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.09097146987915,
+      "learning_rate": 0.0005043,
+      "loss": 1.9936,
       "step": 2662
     },
     {
       "epoch": 4.2608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7409591674804688,
+      "learning_rate": 0.00050415,
+      "loss": 1.7158,
       "step": 2663
     },
     {
       "epoch": 4.2624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9653496742248535,
+      "learning_rate": 0.0005039999999999999,
+      "loss": 1.6247,
       "step": 2664
     },
     {
       "epoch": 4.264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4377232789993286,
+      "learning_rate": 0.0005038499999999999,
+      "loss": 1.796,
       "step": 2665
     },
     {
       "epoch": 4.2656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.103083610534668,
+      "learning_rate": 0.0005036999999999999,
+      "loss": 1.7619,
       "step": 2666
     },
     {
       "epoch": 4.2672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7207717895507812,
+      "learning_rate": 0.00050355,
+      "loss": 1.7422,
       "step": 2667
     },
     {
       "epoch": 4.2688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.597609043121338,
+      "learning_rate": 0.0005034,
+      "loss": 2.0255,
       "step": 2668
     },
     {
       "epoch": 4.2704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.644750118255615,
+      "learning_rate": 0.0005032499999999999,
+      "loss": 1.7337,
       "step": 2669
     },
     {
       "epoch": 4.272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.096191883087158,
+      "learning_rate": 0.0005030999999999999,
+      "loss": 1.7898,
       "step": 2670
     },
     {
       "epoch": 4.2736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.152255058288574,
+      "learning_rate": 0.00050295,
+      "loss": 1.6478,
       "step": 2671
     },
     {
       "epoch": 4.2752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.358015537261963,
+      "learning_rate": 0.0005028,
+      "loss": 1.6115,
       "step": 2672
     },
     {
       "epoch": 4.2768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.264955759048462,
+      "learning_rate": 0.00050265,
+      "loss": 1.8123,
       "step": 2673
     },
     {
       "epoch": 4.2783999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8399302959442139,
+      "learning_rate": 0.0005025,
+      "loss": 1.97,
       "step": 2674
     },
     {
       "epoch": 4.28,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8927841186523438,
+      "learning_rate": 0.0005023499999999999,
+      "loss": 1.7362,
       "step": 2675
     },
     {
       "epoch": 4.2816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1136553287506104,
+      "learning_rate": 0.0005022,
+      "loss": 1.5103,
       "step": 2676
     },
     {
       "epoch": 4.2832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.200005054473877,
+      "learning_rate": 0.00050205,
+      "loss": 1.8307,
       "step": 2677
     },
     {
       "epoch": 4.2848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8706302642822266,
+      "learning_rate": 0.0005019,
+      "loss": 2.0445,
       "step": 2678
     },
     {
       "epoch": 4.2864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9827845096588135,
+      "learning_rate": 0.00050175,
+      "loss": 1.6169,
       "step": 2679
     },
     {
       "epoch": 4.288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5512272119522095,
+      "learning_rate": 0.0005015999999999999,
+      "loss": 2.0119,
       "step": 2680
     },
     {
       "epoch": 4.2896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1824843883514404,
+      "learning_rate": 0.00050145,
+      "loss": 2.1293,
       "step": 2681
     },
     {
       "epoch": 4.2912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.637053370475769,
+      "learning_rate": 0.0005013,
+      "loss": 1.7512,
       "step": 2682
     },
     {
       "epoch": 4.2928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.891478955745697,
+      "learning_rate": 0.00050115,
+      "loss": 1.6766,
       "step": 2683
     },
     {
       "epoch": 4.2943999999999996,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.968799352645874,
+      "learning_rate": 0.0005009999999999999,
+      "loss": 1.7987,
       "step": 2684
     },
     {
       "epoch": 4.296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8135929107666016,
+      "learning_rate": 0.00050085,
+      "loss": 1.513,
       "step": 2685
     },
     {
       "epoch": 4.2976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.973942995071411,
+      "learning_rate": 0.0005007,
+      "loss": 1.9517,
       "step": 2686
     },
     {
       "epoch": 4.2992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5238656997680664,
+      "learning_rate": 0.00050055,
+      "loss": 1.871,
       "step": 2687
     },
     {
       "epoch": 4.3008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8395425081253052,
+      "learning_rate": 0.0005003999999999999,
+      "loss": 2.3133,
       "step": 2688
     },
     {
       "epoch": 4.3024000000000004,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2735295295715332,
+      "learning_rate": 0.0005002499999999999,
+      "loss": 2.0122,
       "step": 2689
     },
     {
       "epoch": 4.304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.705003023147583,
+      "learning_rate": 0.0005001,
+      "loss": 1.7076,
       "step": 2690
     },
     {
       "epoch": 4.3056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0689762830734253,
+      "learning_rate": 0.00049995,
+      "loss": 1.8977,
       "step": 2691
     },
     {
       "epoch": 4.3072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1098942756652832,
+      "learning_rate": 0.0004997999999999999,
+      "loss": 2.1884,
       "step": 2692
     },
     {
       "epoch": 4.3088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3457235097885132,
+      "learning_rate": 0.0004996499999999999,
+      "loss": 1.9898,
       "step": 2693
     },
     {
       "epoch": 4.3104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3474767208099365,
+      "learning_rate": 0.0004994999999999999,
+      "loss": 2.1797,
       "step": 2694
     },
     {
       "epoch": 4.312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.194025754928589,
+      "learning_rate": 0.00049935,
+      "loss": 1.7558,
       "step": 2695
     },
     {
       "epoch": 4.3136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.322290420532227,
+      "learning_rate": 0.0004991999999999999,
+      "loss": 2.1153,
       "step": 2696
     },
     {
       "epoch": 4.3152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1777799129486084,
+      "learning_rate": 0.0004990499999999999,
+      "loss": 1.9527,
       "step": 2697
     },
     {
       "epoch": 4.3168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.741779088973999,
+      "learning_rate": 0.0004988999999999999,
+      "loss": 2.2402,
       "step": 2698
     },
     {
       "epoch": 4.3184000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0727362632751465,
+      "learning_rate": 0.00049875,
+      "loss": 2.1972,
       "step": 2699
     },
     {
       "epoch": 4.32,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2700
-    },
-    {
-      "epoch": 4.32,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 159.7893,
-      "eval_samples_per_second": 19.626,
-      "eval_steps_per_second": 1.227,
-      "eval_wer": 1.0,
+      "grad_norm": 2.3544955253601074,
+      "learning_rate": 0.0004986,
+      "loss": 2.1574,
       "step": 2700
     },
     {
       "epoch": 4.3216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.276401996612549,
+      "learning_rate": 0.00049845,
+      "loss": 2.4749,
       "step": 2701
     },
     {
       "epoch": 4.3232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 12.329696655273438,
+      "learning_rate": 0.0004982999999999999,
+      "loss": 3.0003,
       "step": 2702
     },
     {
       "epoch": 4.3248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4918222427368164,
+      "learning_rate": 0.0004981499999999999,
+      "loss": 2.1861,
       "step": 2703
     },
     {
       "epoch": 4.3264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.620767116546631,
+      "learning_rate": 0.000498,
+      "loss": 2.1404,
       "step": 2704
     },
     {
       "epoch": 4.328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1579036712646484,
+      "learning_rate": 0.00049785,
+      "loss": 1.9692,
       "step": 2705
     },
     {
       "epoch": 4.3296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.295342206954956,
+      "learning_rate": 0.0004977,
+      "loss": 1.9675,
       "step": 2706
     },
     {
       "epoch": 4.3312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0219806432724,
+      "learning_rate": 0.00049755,
+      "loss": 1.7226,
       "step": 2707
     },
     {
       "epoch": 4.3328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3120267391204834,
+      "learning_rate": 0.0004974,
+      "loss": 2.0358,
       "step": 2708
     },
     {
       "epoch": 4.3344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3944854736328125,
+      "learning_rate": 0.00049725,
+      "loss": 2.0762,
       "step": 2709
     },
     {
       "epoch": 4.336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1443352699279785,
+      "learning_rate": 0.0004971,
+      "loss": 1.9482,
       "step": 2710
     },
     {
       "epoch": 4.3376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1645523309707642,
+      "learning_rate": 0.00049695,
+      "loss": 1.7209,
       "step": 2711
     },
     {
       "epoch": 4.3392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0434651374816895,
+      "learning_rate": 0.0004967999999999999,
+      "loss": 2.124,
       "step": 2712
     },
     {
       "epoch": 4.3408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5835245847702026,
+      "learning_rate": 0.00049665,
+      "loss": 1.9263,
       "step": 2713
     },
     {
       "epoch": 4.3424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5039459466934204,
+      "learning_rate": 0.0004965,
+      "loss": 1.9301,
       "step": 2714
     },
     {
       "epoch": 4.344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2301478385925293,
+      "learning_rate": 0.00049635,
+      "loss": 1.8303,
       "step": 2715
     },
     {
       "epoch": 4.3456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2232022285461426,
+      "learning_rate": 0.0004961999999999999,
+      "loss": 1.9568,
       "step": 2716
     },
     {
       "epoch": 4.3472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9055202603340149,
+      "learning_rate": 0.0004960499999999999,
+      "loss": 1.6978,
       "step": 2717
     },
     {
       "epoch": 4.3488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9296348690986633,
+      "learning_rate": 0.0004959,
+      "loss": 1.5963,
       "step": 2718
     },
     {
       "epoch": 4.3504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9808449745178223,
+      "learning_rate": 0.00049575,
+      "loss": 1.5989,
       "step": 2719
     },
     {
       "epoch": 4.352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.236293315887451,
+      "learning_rate": 0.0004955999999999999,
+      "loss": 1.9088,
       "step": 2720
     },
     {
       "epoch": 4.3536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9127609729766846,
+      "learning_rate": 0.0004954499999999999,
+      "loss": 1.6453,
       "step": 2721
     },
     {
       "epoch": 4.3552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.41826868057251,
+      "learning_rate": 0.0004953,
+      "loss": 1.9933,
       "step": 2722
     },
     {
       "epoch": 4.3568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.878091812133789,
+      "learning_rate": 0.00049515,
+      "loss": 1.7266,
       "step": 2723
     },
     {
       "epoch": 4.3584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5814834833145142,
+      "learning_rate": 0.0004949999999999999,
+      "loss": 1.7923,
       "step": 2724
     },
     {
       "epoch": 4.36,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.397496461868286,
+      "learning_rate": 0.0004948499999999999,
+      "loss": 1.7079,
       "step": 2725
     },
     {
       "epoch": 4.3616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9779950380325317,
+      "learning_rate": 0.0004946999999999999,
+      "loss": 1.9622,
       "step": 2726
     },
     {
       "epoch": 4.3632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6545337438583374,
+      "learning_rate": 0.00049455,
+      "loss": 2.1439,
       "step": 2727
     },
     {
       "epoch": 4.3648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8302526473999023,
+      "learning_rate": 0.0004944,
+      "loss": 2.0157,
       "step": 2728
     },
     {
       "epoch": 4.3664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.294466257095337,
+      "learning_rate": 0.0004942499999999999,
+      "loss": 1.6352,
       "step": 2729
     },
     {
       "epoch": 4.368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.946574091911316,
+      "learning_rate": 0.0004940999999999999,
+      "loss": 1.7012,
       "step": 2730
     },
     {
       "epoch": 4.3696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.769045889377594,
+      "learning_rate": 0.0004939499999999999,
+      "loss": 1.6464,
       "step": 2731
     },
     {
       "epoch": 4.3712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6270204782485962,
+      "learning_rate": 0.0004938,
+      "loss": 1.9476,
       "step": 2732
     },
     {
       "epoch": 4.3728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.318570137023926,
+      "learning_rate": 0.00049365,
+      "loss": 1.7405,
       "step": 2733
     },
     {
       "epoch": 4.3744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.380737066268921,
+      "learning_rate": 0.0004935,
+      "loss": 1.8886,
       "step": 2734
     },
     {
       "epoch": 4.376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8267993927001953,
+      "learning_rate": 0.0004933499999999999,
+      "loss": 1.7565,
       "step": 2735
     },
     {
       "epoch": 4.3776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3764233589172363,
+      "learning_rate": 0.0004932,
+      "loss": 1.8319,
       "step": 2736
     },
     {
       "epoch": 4.3792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.615099549293518,
+      "learning_rate": 0.00049305,
+      "loss": 1.5021,
       "step": 2737
     },
     {
       "epoch": 4.3808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2117981910705566,
+      "learning_rate": 0.0004929,
+      "loss": 1.7998,
       "step": 2738
     },
     {
       "epoch": 4.3824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9510959386825562,
+      "learning_rate": 0.00049275,
+      "loss": 1.9194,
       "step": 2739
     },
     {
       "epoch": 4.384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5256377458572388,
+      "learning_rate": 0.0004925999999999999,
+      "loss": 2.4345,
       "step": 2740
     },
     {
       "epoch": 4.3856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9747059941291809,
+      "learning_rate": 0.00049245,
+      "loss": 2.1591,
       "step": 2741
     },
     {
       "epoch": 4.3872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8639597296714783,
+      "learning_rate": 0.0004923,
+      "loss": 1.7427,
       "step": 2742
     },
     {
       "epoch": 4.3888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0038577318191528,
+      "learning_rate": 0.00049215,
+      "loss": 1.7629,
       "step": 2743
     },
     {
       "epoch": 4.3904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7393220663070679,
+      "learning_rate": 0.0004919999999999999,
+      "loss": 2.0587,
       "step": 2744
     },
     {
       "epoch": 4.392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.120962619781494,
+      "learning_rate": 0.00049185,
+      "loss": 1.7976,
       "step": 2745
     },
     {
       "epoch": 4.3936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9911690950393677,
+      "learning_rate": 0.0004917,
+      "loss": 2.1529,
       "step": 2746
     },
     {
       "epoch": 4.3952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.39595365524292,
+      "learning_rate": 0.00049155,
+      "loss": 2.2529,
       "step": 2747
     },
     {
       "epoch": 4.3968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0449539422988892,
+      "learning_rate": 0.0004913999999999999,
+      "loss": 2.2485,
       "step": 2748
     },
     {
       "epoch": 4.3984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2274101972579956,
+      "learning_rate": 0.0004912499999999999,
+      "loss": 2.6684,
       "step": 2749
     },
     {
       "epoch": 4.4,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6110919713974,
+      "learning_rate": 0.0004911,
+      "loss": 2.3316,
       "step": 2750
     },
     {
       "epoch": 4.4016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 31.37392234802246,
+      "learning_rate": 0.00049095,
+      "loss": 3.3606,
       "step": 2751
     },
     {
       "epoch": 4.4032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.821800231933594,
+      "learning_rate": 0.0004907999999999999,
+      "loss": 2.6849,
       "step": 2752
     },
     {
       "epoch": 4.4048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9470937252044678,
+      "learning_rate": 0.0004906499999999999,
+      "loss": 2.1471,
       "step": 2753
     },
     {
       "epoch": 4.4064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6643931865692139,
+      "learning_rate": 0.0004904999999999999,
+      "loss": 2.1997,
       "step": 2754
     },
     {
       "epoch": 4.408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0974528789520264,
+      "learning_rate": 0.00049035,
+      "loss": 2.2333,
       "step": 2755
     },
     {
       "epoch": 4.4096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.75770092010498,
+      "learning_rate": 0.0004901999999999999,
+      "loss": 2.7282,
       "step": 2756
     },
     {
       "epoch": 4.4112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.481705665588379,
+      "learning_rate": 0.0004900499999999999,
+      "loss": 2.1102,
       "step": 2757
     },
     {
       "epoch": 4.4128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0984652042388916,
+      "learning_rate": 0.0004898999999999999,
+      "loss": 2.0301,
       "step": 2758
     },
     {
       "epoch": 4.4144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.651517391204834,
+      "learning_rate": 0.00048975,
+      "loss": 1.9718,
       "step": 2759
     },
     {
       "epoch": 4.416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.137758493423462,
+      "learning_rate": 0.0004896,
+      "loss": 1.8473,
       "step": 2760
     },
     {
       "epoch": 4.4176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6671268939971924,
+      "learning_rate": 0.0004894499999999999,
+      "loss": 2.1645,
       "step": 2761
     },
     {
       "epoch": 4.4192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.523720383644104,
+      "learning_rate": 0.0004892999999999999,
+      "loss": 1.8029,
       "step": 2762
     },
     {
       "epoch": 4.4208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.891670227050781,
+      "learning_rate": 0.0004891499999999999,
+      "loss": 2.1067,
       "step": 2763
     },
     {
       "epoch": 4.4224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.942965030670166,
+      "learning_rate": 0.000489,
+      "loss": 1.6909,
       "step": 2764
     },
     {
       "epoch": 4.424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.303746223449707,
+      "learning_rate": 0.00048885,
+      "loss": 2.1012,
       "step": 2765
     },
     {
       "epoch": 4.4256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8003095388412476,
+      "learning_rate": 0.0004887,
+      "loss": 2.0751,
       "step": 2766
     },
     {
       "epoch": 4.4272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.143497943878174,
+      "learning_rate": 0.0004885499999999999,
+      "loss": 1.866,
       "step": 2767
     },
     {
       "epoch": 4.4288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6083755493164062,
+      "learning_rate": 0.0004883999999999999,
+      "loss": 1.8097,
       "step": 2768
     },
     {
       "epoch": 4.4304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3343874216079712,
+      "learning_rate": 0.0004882499999999999,
+      "loss": 1.8029,
       "step": 2769
     },
     {
       "epoch": 4.432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.8573079109191895,
+      "learning_rate": 0.00048809999999999994,
+      "loss": 1.7581,
       "step": 2770
     },
     {
       "epoch": 4.4336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.798096656799316,
+      "learning_rate": 0.00048794999999999996,
+      "loss": 2.0261,
       "step": 2771
     },
     {
       "epoch": 4.4352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.447819232940674,
+      "learning_rate": 0.00048779999999999993,
+      "loss": 1.6978,
       "step": 2772
     },
     {
       "epoch": 4.4368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9434682130813599,
+      "learning_rate": 0.00048764999999999995,
+      "loss": 1.5958,
       "step": 2773
     },
     {
       "epoch": 4.4384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6188219785690308,
+      "learning_rate": 0.0004875,
+      "loss": 2.3481,
       "step": 2774
     },
     {
       "epoch": 4.44,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8848549127578735,
+      "learning_rate": 0.00048734999999999995,
+      "loss": 1.8844,
       "step": 2775
     },
     {
       "epoch": 4.4416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.049158811569214,
+      "learning_rate": 0.00048719999999999997,
+      "loss": 1.8728,
       "step": 2776
     },
     {
       "epoch": 4.4432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0514065027236938,
+      "learning_rate": 0.00048704999999999994,
+      "loss": 1.8622,
       "step": 2777
     },
     {
       "epoch": 4.4448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5144424438476562,
+      "learning_rate": 0.00048689999999999996,
+      "loss": 1.887,
       "step": 2778
     },
     {
       "epoch": 4.4464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3987104892730713,
+      "learning_rate": 0.00048675,
+      "loss": 1.8795,
       "step": 2779
     },
     {
       "epoch": 4.448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4233217239379883,
+      "learning_rate": 0.0004866,
+      "loss": 1.9542,
       "step": 2780
     },
     {
       "epoch": 4.4496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9357428550720215,
+      "learning_rate": 0.0004864499999999999,
+      "loss": 1.8799,
       "step": 2781
     },
     {
       "epoch": 4.4512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.104844093322754,
+      "learning_rate": 0.00048629999999999995,
+      "loss": 1.6141,
       "step": 2782
     },
     {
       "epoch": 4.4528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3217804431915283,
+      "learning_rate": 0.00048614999999999997,
+      "loss": 1.7236,
       "step": 2783
     },
     {
       "epoch": 4.4544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2796238660812378,
+      "learning_rate": 0.000486,
+      "loss": 2.0714,
       "step": 2784
     },
     {
       "epoch": 4.456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0485546588897705,
+      "learning_rate": 0.00048584999999999996,
+      "loss": 1.8654,
       "step": 2785
     },
     {
       "epoch": 4.4576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.945281028747559,
+      "learning_rate": 0.00048569999999999993,
+      "loss": 1.7194,
       "step": 2786
     },
     {
       "epoch": 4.4592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.205522298812866,
+      "learning_rate": 0.00048554999999999996,
+      "loss": 1.8191,
       "step": 2787
     },
     {
       "epoch": 4.4608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.893653392791748,
+      "learning_rate": 0.0004854,
+      "loss": 1.9642,
       "step": 2788
     },
     {
       "epoch": 4.4624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2910706996917725,
+      "learning_rate": 0.00048524999999999995,
+      "loss": 2.0733,
       "step": 2789
     },
     {
       "epoch": 4.464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.876647710800171,
+      "learning_rate": 0.0004851,
+      "loss": 2.0299,
       "step": 2790
     },
     {
       "epoch": 4.4656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0618470907211304,
+      "learning_rate": 0.00048495,
+      "loss": 1.8538,
       "step": 2791
     },
     {
       "epoch": 4.4672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4978771209716797,
+      "learning_rate": 0.00048479999999999997,
+      "loss": 1.8727,
       "step": 2792
     },
     {
       "epoch": 4.4688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.992602825164795,
+      "learning_rate": 0.00048464999999999994,
+      "loss": 2.0375,
       "step": 2793
     },
     {
       "epoch": 4.4704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.787044525146484,
+      "learning_rate": 0.00048449999999999996,
+      "loss": 2.1719,
       "step": 2794
     },
     {
       "epoch": 4.4719999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2934472560882568,
+      "learning_rate": 0.00048435,
+      "loss": 1.6233,
       "step": 2795
     },
     {
       "epoch": 4.4736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5270168781280518,
+      "learning_rate": 0.0004842,
+      "loss": 2.1265,
       "step": 2796
     },
     {
       "epoch": 4.4752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.369183540344238,
+      "learning_rate": 0.0004840499999999999,
+      "loss": 1.9866,
       "step": 2797
     },
     {
       "epoch": 4.4768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1386245489120483,
+      "learning_rate": 0.00048389999999999994,
+      "loss": 2.4311,
       "step": 2798
     },
     {
       "epoch": 4.4784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0232652425765991,
+      "learning_rate": 0.00048374999999999997,
+      "loss": 1.8705,
       "step": 2799
     },
     {
       "epoch": 4.48,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2800
-    },
-    {
-      "epoch": 4.48,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 158.6637,
-      "eval_samples_per_second": 19.765,
-      "eval_steps_per_second": 1.235,
-      "eval_wer": 1.0,
+      "grad_norm": 3.219254970550537,
+      "learning_rate": 0.0004836,
+      "loss": 2.4248,
       "step": 2800
     },
     {
       "epoch": 4.4816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 20.291139602661133,
+      "learning_rate": 0.00048344999999999996,
+      "loss": 3.2457,
       "step": 2801
     },
     {
       "epoch": 4.4832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5292768478393555,
+      "learning_rate": 0.00048329999999999993,
+      "loss": 1.8494,
       "step": 2802
     },
     {
       "epoch": 4.4848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 16.353145599365234,
+      "learning_rate": 0.00048314999999999995,
+      "loss": 2.4255,
       "step": 2803
     },
     {
       "epoch": 4.4864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 27.58832550048828,
+      "learning_rate": 0.000483,
+      "loss": 3.1663,
       "step": 2804
     },
     {
       "epoch": 4.4879999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.299684524536133,
+      "learning_rate": 0.00048284999999999995,
+      "loss": 2.4567,
       "step": 2805
     },
     {
       "epoch": 4.4896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7438368797302246,
+      "learning_rate": 0.00048269999999999997,
+      "loss": 2.0492,
       "step": 2806
     },
     {
       "epoch": 4.4912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.070641040802002,
+      "learning_rate": 0.00048255,
+      "loss": 2.2044,
       "step": 2807
     },
     {
       "epoch": 4.4928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.334407925605774,
+      "learning_rate": 0.00048239999999999996,
+      "loss": 1.9938,
       "step": 2808
     },
     {
       "epoch": 4.4944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9917300939559937,
+      "learning_rate": 0.00048224999999999993,
+      "loss": 1.8998,
       "step": 2809
     },
     {
       "epoch": 4.496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3776772022247314,
+      "learning_rate": 0.00048209999999999995,
+      "loss": 1.8164,
       "step": 2810
     },
     {
       "epoch": 4.4976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.257997751235962,
+      "learning_rate": 0.00048195,
+      "loss": 1.8108,
       "step": 2811
     },
     {
       "epoch": 4.4992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6873910427093506,
+      "learning_rate": 0.0004818,
+      "loss": 2.3107,
       "step": 2812
     },
     {
       "epoch": 4.5008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2140886783599854,
+      "learning_rate": 0.0004816499999999999,
+      "loss": 1.6474,
       "step": 2813
     },
     {
       "epoch": 4.5024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3614856004714966,
+      "learning_rate": 0.00048149999999999994,
+      "loss": 1.9323,
       "step": 2814
     },
     {
       "epoch": 4.504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0496727228164673,
+      "learning_rate": 0.00048134999999999996,
+      "loss": 1.6896,
       "step": 2815
     },
     {
       "epoch": 4.5056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8492391109466553,
+      "learning_rate": 0.0004812,
+      "loss": 1.748,
       "step": 2816
     },
     {
       "epoch": 4.5072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.207594394683838,
+      "learning_rate": 0.00048104999999999996,
+      "loss": 1.6843,
       "step": 2817
     },
     {
       "epoch": 4.5088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5215344429016113,
+      "learning_rate": 0.0004808999999999999,
+      "loss": 1.7607,
       "step": 2818
     },
     {
       "epoch": 4.5104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.886686325073242,
+      "learning_rate": 0.00048074999999999995,
+      "loss": 2.1518,
       "step": 2819
     },
     {
       "epoch": 4.5120000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.15547513961792,
+      "learning_rate": 0.00048059999999999997,
+      "loss": 2.1099,
       "step": 2820
     },
     {
       "epoch": 4.5136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1715599298477173,
+      "learning_rate": 0.00048044999999999994,
+      "loss": 1.8574,
       "step": 2821
     },
     {
       "epoch": 4.5152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2730066776275635,
+      "learning_rate": 0.00048029999999999997,
+      "loss": 2.2673,
       "step": 2822
     },
     {
       "epoch": 4.5168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0949064493179321,
+      "learning_rate": 0.00048015,
+      "loss": 1.7473,
       "step": 2823
     },
     {
       "epoch": 4.5184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.350168228149414,
+      "learning_rate": 0.00047999999999999996,
+      "loss": 1.712,
       "step": 2824
     },
     {
       "epoch": 4.52,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9208452701568604,
+      "learning_rate": 0.00047984999999999993,
+      "loss": 1.7704,
       "step": 2825
     },
     {
       "epoch": 4.5216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5936332941055298,
+      "learning_rate": 0.00047969999999999995,
+      "loss": 2.0433,
       "step": 2826
     },
     {
       "epoch": 4.5232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.146118402481079,
+      "learning_rate": 0.00047955,
+      "loss": 1.9635,
       "step": 2827
     },
     {
       "epoch": 4.5248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.349436044692993,
+      "learning_rate": 0.0004794,
+      "loss": 1.8187,
       "step": 2828
     },
     {
       "epoch": 4.5264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8291957378387451,
+      "learning_rate": 0.0004792499999999999,
+      "loss": 1.8534,
       "step": 2829
     },
     {
       "epoch": 4.5280000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6717575788497925,
+      "learning_rate": 0.00047909999999999994,
+      "loss": 1.7578,
       "step": 2830
     },
     {
       "epoch": 4.5296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.676916241645813,
+      "learning_rate": 0.00047894999999999996,
+      "loss": 2.0669,
       "step": 2831
     },
     {
       "epoch": 4.5312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7795640230178833,
+      "learning_rate": 0.0004788,
+      "loss": 1.8332,
       "step": 2832
     },
     {
       "epoch": 4.5328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9749438762664795,
+      "learning_rate": 0.00047864999999999995,
+      "loss": 1.8182,
       "step": 2833
     },
     {
       "epoch": 4.5344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5495363473892212,
+      "learning_rate": 0.0004785,
+      "loss": 2.0464,
       "step": 2834
     },
     {
       "epoch": 4.536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3625857830047607,
+      "learning_rate": 0.00047834999999999994,
+      "loss": 2.1856,
       "step": 2835
     },
     {
       "epoch": 4.5376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.267985820770264,
+      "learning_rate": 0.00047819999999999997,
+      "loss": 1.5406,
       "step": 2836
     },
     {
       "epoch": 4.5392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.528249263763428,
+      "learning_rate": 0.00047804999999999994,
+      "loss": 2.138,
       "step": 2837
     },
     {
       "epoch": 4.5408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.518407106399536,
+      "learning_rate": 0.00047789999999999996,
+      "loss": 1.7701,
       "step": 2838
     },
     {
       "epoch": 4.5424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.524309158325195,
+      "learning_rate": 0.00047775,
+      "loss": 1.8426,
       "step": 2839
     },
     {
       "epoch": 4.5440000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2349860668182373,
+      "learning_rate": 0.0004776,
+      "loss": 1.7634,
       "step": 2840
     },
     {
       "epoch": 4.5456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0253136157989502,
+      "learning_rate": 0.0004774499999999999,
+      "loss": 1.9,
       "step": 2841
     },
     {
       "epoch": 4.5472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.908934593200684,
+      "learning_rate": 0.00047729999999999995,
+      "loss": 2.2211,
       "step": 2842
     },
     {
       "epoch": 4.5488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.103787899017334,
+      "learning_rate": 0.00047714999999999997,
+      "loss": 1.7664,
       "step": 2843
     },
     {
       "epoch": 4.5504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5886802673339844,
+      "learning_rate": 0.000477,
+      "loss": 2.1891,
       "step": 2844
     },
     {
       "epoch": 4.552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.039703845977783,
+      "learning_rate": 0.0004768499999999999,
+      "loss": 2.2356,
       "step": 2845
     },
     {
       "epoch": 4.5536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9848023653030396,
+      "learning_rate": 0.00047669999999999993,
+      "loss": 1.7484,
       "step": 2846
     },
     {
       "epoch": 4.5552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5086679458618164,
+      "learning_rate": 0.00047654999999999996,
+      "loss": 2.4515,
       "step": 2847
     },
     {
       "epoch": 4.5568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0788493156433105,
+      "learning_rate": 0.0004764,
+      "loss": 2.3566,
       "step": 2848
     },
     {
       "epoch": 4.5584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1540278196334839,
+      "learning_rate": 0.00047624999999999995,
+      "loss": 2.5733,
       "step": 2849
     },
     {
       "epoch": 4.5600000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1869611740112305,
+      "learning_rate": 0.00047609999999999997,
+      "loss": 2.7406,
       "step": 2850
     },
     {
       "epoch": 4.5616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 18.052366256713867,
+      "learning_rate": 0.00047594999999999994,
+      "loss": 2.3533,
       "step": 2851
     },
     {
       "epoch": 4.5632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 12.693831443786621,
+      "learning_rate": 0.00047579999999999996,
+      "loss": 2.6903,
       "step": 2852
     },
     {
       "epoch": 4.5648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 12.301849365234375,
+      "learning_rate": 0.00047564999999999993,
+      "loss": 2.5039,
       "step": 2853
     },
     {
       "epoch": 4.5664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.84140682220459,
+      "learning_rate": 0.00047549999999999996,
+      "loss": 2.4382,
       "step": 2854
     },
     {
       "epoch": 4.568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.18992805480957,
+      "learning_rate": 0.00047535,
+      "loss": 2.265,
       "step": 2855
     },
     {
       "epoch": 4.5696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.735015392303467,
+      "learning_rate": 0.0004752,
+      "loss": 1.8568,
       "step": 2856
     },
     {
       "epoch": 4.5712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.784823179244995,
+      "learning_rate": 0.0004750499999999999,
+      "loss": 2.0006,
       "step": 2857
     },
     {
       "epoch": 4.5728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5699920654296875,
+      "learning_rate": 0.00047489999999999994,
+      "loss": 1.8837,
       "step": 2858
     },
     {
       "epoch": 4.5744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1189074516296387,
+      "learning_rate": 0.00047474999999999997,
+      "loss": 1.9481,
       "step": 2859
     },
     {
       "epoch": 4.576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3212521076202393,
+      "learning_rate": 0.0004746,
+      "loss": 1.5919,
       "step": 2860
     },
     {
       "epoch": 4.5776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.494596481323242,
+      "learning_rate": 0.0004744499999999999,
+      "loss": 2.3915,
       "step": 2861
     },
     {
       "epoch": 4.5792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6995341777801514,
+      "learning_rate": 0.00047429999999999993,
+      "loss": 1.9911,
       "step": 2862
     },
     {
       "epoch": 4.5808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6684291362762451,
+      "learning_rate": 0.00047414999999999995,
+      "loss": 2.093,
       "step": 2863
     },
     {
       "epoch": 4.5824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3050556182861328,
+      "learning_rate": 0.000474,
+      "loss": 1.8913,
       "step": 2864
     },
     {
       "epoch": 4.584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.187692880630493,
+      "learning_rate": 0.00047384999999999994,
+      "loss": 1.8978,
       "step": 2865
     },
     {
       "epoch": 4.5856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.71722412109375,
+      "learning_rate": 0.00047369999999999997,
+      "loss": 1.9785,
       "step": 2866
     },
     {
       "epoch": 4.5872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2038593292236328,
+      "learning_rate": 0.00047354999999999994,
+      "loss": 1.5634,
       "step": 2867
     },
     {
       "epoch": 4.5888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5433554649353027,
+      "learning_rate": 0.00047339999999999996,
+      "loss": 1.8322,
       "step": 2868
     },
     {
       "epoch": 4.5904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5715316534042358,
+      "learning_rate": 0.00047324999999999993,
+      "loss": 1.8705,
       "step": 2869
     },
     {
       "epoch": 4.592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.9844120144844055,
+      "learning_rate": 0.00047309999999999995,
+      "loss": 1.8422,
       "step": 2870
     },
     {
       "epoch": 4.5936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.68367338180542,
+      "learning_rate": 0.00047295,
+      "loss": 2.1635,
       "step": 2871
     },
     {
       "epoch": 4.5952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4707579612731934,
+      "learning_rate": 0.0004728,
+      "loss": 1.8738,
       "step": 2872
     },
     {
       "epoch": 4.5968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1058239936828613,
+      "learning_rate": 0.0004726499999999999,
+      "loss": 1.7108,
       "step": 2873
     },
     {
       "epoch": 4.5984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5762336254119873,
+      "learning_rate": 0.00047249999999999994,
+      "loss": 1.4881,
       "step": 2874
     },
     {
       "epoch": 4.6,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.265877366065979,
+      "learning_rate": 0.00047234999999999996,
+      "loss": 2.0167,
       "step": 2875
     },
     {
       "epoch": 4.6016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.916661024093628,
+      "learning_rate": 0.0004722,
+      "loss": 1.6457,
       "step": 2876
     },
     {
       "epoch": 4.6032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.10941219329834,
+      "learning_rate": 0.00047204999999999995,
+      "loss": 1.7151,
       "step": 2877
     },
     {
       "epoch": 4.6048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0412096977233887,
+      "learning_rate": 0.0004718999999999999,
+      "loss": 1.5545,
       "step": 2878
     },
     {
       "epoch": 4.6064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.154681921005249,
+      "learning_rate": 0.00047174999999999995,
+      "loss": 1.5489,
       "step": 2879
     },
     {
       "epoch": 4.608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9898120164871216,
+      "learning_rate": 0.00047159999999999997,
+      "loss": 2.1657,
       "step": 2880
     },
     {
       "epoch": 4.6096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4565964937210083,
+      "learning_rate": 0.00047144999999999994,
+      "loss": 1.453,
       "step": 2881
     },
     {
       "epoch": 4.6112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1306456327438354,
+      "learning_rate": 0.00047129999999999996,
+      "loss": 1.9696,
       "step": 2882
     },
     {
       "epoch": 4.6128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4046789407730103,
+      "learning_rate": 0.00047115,
+      "loss": 1.9966,
       "step": 2883
     },
     {
       "epoch": 4.6144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3461419343948364,
+      "learning_rate": 0.00047099999999999996,
+      "loss": 1.576,
       "step": 2884
     },
     {
       "epoch": 4.616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1443793773651123,
+      "learning_rate": 0.0004708499999999999,
+      "loss": 2.1933,
       "step": 2885
     },
     {
       "epoch": 4.6176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0423054695129395,
+      "learning_rate": 0.00047069999999999995,
+      "loss": 1.9633,
       "step": 2886
     },
     {
       "epoch": 4.6192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0566394329071045,
+      "learning_rate": 0.00047054999999999997,
+      "loss": 1.8235,
       "step": 2887
     },
     {
       "epoch": 4.6208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.431408166885376,
+      "learning_rate": 0.0004704,
+      "loss": 1.7626,
       "step": 2888
     },
     {
       "epoch": 4.6224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2627670764923096,
+      "learning_rate": 0.0004702499999999999,
+      "loss": 2.1956,
       "step": 2889
     },
     {
       "epoch": 4.624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.180410146713257,
+      "learning_rate": 0.00047009999999999993,
+      "loss": 2.3186,
       "step": 2890
     },
     {
       "epoch": 4.6256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5140712261199951,
+      "learning_rate": 0.00046994999999999996,
+      "loss": 2.1441,
       "step": 2891
     },
     {
       "epoch": 4.6272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.021548271179199,
+      "learning_rate": 0.0004698,
+      "loss": 2.1141,
       "step": 2892
     },
     {
       "epoch": 4.6288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0709006786346436,
+      "learning_rate": 0.00046964999999999995,
+      "loss": 1.9173,
       "step": 2893
     },
     {
       "epoch": 4.6304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4512224197387695,
+      "learning_rate": 0.0004694999999999999,
+      "loss": 2.2301,
       "step": 2894
     },
     {
       "epoch": 4.632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.720222234725952,
+      "learning_rate": 0.00046934999999999994,
+      "loss": 2.0691,
       "step": 2895
     },
     {
       "epoch": 4.6336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0958988666534424,
+      "learning_rate": 0.00046919999999999997,
+      "loss": 1.8199,
       "step": 2896
     },
     {
       "epoch": 4.6352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1114755868911743,
+      "learning_rate": 0.00046904999999999994,
+      "loss": 1.9451,
       "step": 2897
     },
     {
       "epoch": 4.6368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1838823556900024,
+      "learning_rate": 0.00046889999999999996,
+      "loss": 2.1416,
       "step": 2898
     },
     {
       "epoch": 4.6384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0702041387557983,
+      "learning_rate": 0.00046875,
+      "loss": 2.1344,
       "step": 2899
     },
     {
       "epoch": 4.64,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 2900
-    },
-    {
-      "epoch": 4.64,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 156.9728,
-      "eval_samples_per_second": 19.978,
-      "eval_steps_per_second": 1.249,
-      "eval_wer": 1.0,
+      "grad_norm": 2.476487398147583,
+      "learning_rate": 0.00046859999999999995,
+      "loss": 2.2368,
       "step": 2900
     },
     {
       "epoch": 4.6416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 26.568328857421875,
+      "learning_rate": 0.00046845,
+      "loss": 3.5976,
       "step": 2901
     },
     {
       "epoch": 4.6432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8742352724075317,
+      "learning_rate": 0.00046829999999999994,
+      "loss": 2.0308,
       "step": 2902
     },
     {
       "epoch": 4.6448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 14.89755630493164,
+      "learning_rate": 0.00046814999999999997,
+      "loss": 2.6192,
       "step": 2903
     },
     {
       "epoch": 4.6464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.758499145507812,
+      "learning_rate": 0.000468,
+      "loss": 2.4896,
       "step": 2904
     },
     {
       "epoch": 4.648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8912417888641357,
+      "learning_rate": 0.00046785,
+      "loss": 2.2466,
       "step": 2905
     },
     {
       "epoch": 4.6495999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 15.772459030151367,
+      "learning_rate": 0.00046769999999999993,
+      "loss": 2.9467,
       "step": 2906
     },
     {
       "epoch": 4.6512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.318552494049072,
+      "learning_rate": 0.00046754999999999995,
+      "loss": 2.4392,
       "step": 2907
     },
     {
       "epoch": 4.6528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.18526029586792,
+      "learning_rate": 0.0004674,
+      "loss": 2.5809,
       "step": 2908
     },
     {
       "epoch": 4.6544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1041102409362793,
+      "learning_rate": 0.00046725,
+      "loss": 2.1863,
       "step": 2909
     },
     {
       "epoch": 4.656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.712016582489014,
+      "learning_rate": 0.0004670999999999999,
+      "loss": 1.8193,
       "step": 2910
     },
     {
       "epoch": 4.6576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.29626727104187,
+      "learning_rate": 0.00046694999999999994,
+      "loss": 1.9008,
       "step": 2911
     },
     {
       "epoch": 4.6592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.208040714263916,
+      "learning_rate": 0.00046679999999999996,
+      "loss": 1.9267,
       "step": 2912
     },
     {
       "epoch": 4.6608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9327380657196045,
+      "learning_rate": 0.00046665,
+      "loss": 1.9868,
       "step": 2913
     },
     {
       "epoch": 4.6624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5247790813446045,
+      "learning_rate": 0.00046649999999999996,
+      "loss": 1.7713,
       "step": 2914
     },
     {
       "epoch": 4.664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.439487934112549,
+      "learning_rate": 0.00046635,
+      "loss": 2.0238,
       "step": 2915
     },
     {
       "epoch": 4.6655999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8292617797851562,
+      "learning_rate": 0.00046619999999999995,
+      "loss": 1.7811,
       "step": 2916
     },
     {
       "epoch": 4.6672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1137208938598633,
+      "learning_rate": 0.00046604999999999997,
+      "loss": 1.8638,
       "step": 2917
     },
     {
       "epoch": 4.6688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3321573734283447,
+      "learning_rate": 0.00046589999999999994,
+      "loss": 1.8959,
       "step": 2918
     },
     {
       "epoch": 4.6704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2265563011169434,
+      "learning_rate": 0.00046574999999999996,
+      "loss": 1.7412,
       "step": 2919
     },
     {
       "epoch": 4.672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4950315952301025,
+      "learning_rate": 0.0004656,
+      "loss": 1.7304,
       "step": 2920
     },
     {
       "epoch": 4.6736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3217532634735107,
+      "learning_rate": 0.00046545,
+      "loss": 1.6422,
       "step": 2921
     },
     {
       "epoch": 4.6752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4180880784988403,
+      "learning_rate": 0.0004652999999999999,
+      "loss": 2.1201,
       "step": 2922
     },
     {
       "epoch": 4.6768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.970060110092163,
+      "learning_rate": 0.00046514999999999995,
+      "loss": 1.7879,
       "step": 2923
     },
     {
       "epoch": 4.6784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.532371997833252,
+      "learning_rate": 0.00046499999999999997,
+      "loss": 2.1768,
       "step": 2924
     },
     {
       "epoch": 4.68,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.154553174972534,
+      "learning_rate": 0.00046485,
+      "loss": 1.8605,
       "step": 2925
     },
     {
       "epoch": 4.6815999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9632632732391357,
+      "learning_rate": 0.00046469999999999997,
+      "loss": 2.1543,
       "step": 2926
     },
     {
       "epoch": 4.6832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.718745708465576,
+      "learning_rate": 0.00046454999999999993,
+      "loss": 1.7158,
       "step": 2927
     },
     {
       "epoch": 4.6848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.1586384773254395,
+      "learning_rate": 0.00046439999999999996,
+      "loss": 1.8619,
       "step": 2928
     },
     {
       "epoch": 4.6864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7387466430664062,
+      "learning_rate": 0.00046425,
+      "loss": 1.7637,
       "step": 2929
     },
     {
       "epoch": 4.688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1410984992980957,
+      "learning_rate": 0.00046409999999999995,
+      "loss": 1.689,
       "step": 2930
     },
     {
       "epoch": 4.6896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7668966054916382,
+      "learning_rate": 0.00046395,
+      "loss": 2.3125,
       "step": 2931
     },
     {
       "epoch": 4.6912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.661885976791382,
+      "learning_rate": 0.0004638,
+      "loss": 1.8615,
       "step": 2932
     },
     {
       "epoch": 4.6928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0198774337768555,
+      "learning_rate": 0.00046364999999999997,
+      "loss": 2.1258,
       "step": 2933
     },
     {
       "epoch": 4.6944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6803481578826904,
+      "learning_rate": 0.00046349999999999994,
+      "loss": 1.838,
       "step": 2934
     },
     {
       "epoch": 4.696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.10424542427063,
+      "learning_rate": 0.00046334999999999996,
+      "loss": 2.2768,
       "step": 2935
     },
     {
       "epoch": 4.6975999999999996,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8201396465301514,
+      "learning_rate": 0.0004632,
+      "loss": 1.6918,
       "step": 2936
     },
     {
       "epoch": 4.6992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3550063371658325,
+      "learning_rate": 0.00046305,
+      "loss": 1.9115,
       "step": 2937
     },
     {
       "epoch": 4.7008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.885533332824707,
+      "learning_rate": 0.0004628999999999999,
+      "loss": 1.8377,
       "step": 2938
     },
     {
       "epoch": 4.7024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.513812065124512,
+      "learning_rate": 0.00046274999999999995,
+      "loss": 2.2139,
       "step": 2939
     },
     {
       "epoch": 4.704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.13215970993042,
+      "learning_rate": 0.00046259999999999997,
+      "loss": 2.3479,
       "step": 2940
     },
     {
       "epoch": 4.7056000000000004,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0954887866973877,
+      "learning_rate": 0.00046245,
+      "loss": 1.9406,
       "step": 2941
     },
     {
       "epoch": 4.7072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6229615211486816,
+      "learning_rate": 0.00046229999999999996,
+      "loss": 1.9179,
       "step": 2942
     },
     {
       "epoch": 4.7088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.992983341217041,
+      "learning_rate": 0.00046214999999999993,
+      "loss": 1.7419,
       "step": 2943
     },
     {
       "epoch": 4.7104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7511937618255615,
+      "learning_rate": 0.00046199999999999995,
+      "loss": 2.2057,
       "step": 2944
     },
     {
       "epoch": 4.712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7940322160720825,
+      "learning_rate": 0.00046185,
+      "loss": 2.1836,
       "step": 2945
     },
     {
       "epoch": 4.7136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.443817615509033,
+      "learning_rate": 0.00046169999999999995,
+      "loss": 1.7678,
       "step": 2946
     },
     {
       "epoch": 4.7152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4557013511657715,
+      "learning_rate": 0.00046154999999999997,
+      "loss": 1.7645,
       "step": 2947
     },
     {
       "epoch": 4.7168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7131155729293823,
+      "learning_rate": 0.0004614,
+      "loss": 1.9185,
       "step": 2948
     },
     {
       "epoch": 4.7184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7961368560791016,
+      "learning_rate": 0.00046124999999999996,
+      "loss": 2.414,
       "step": 2949
     },
     {
       "epoch": 4.72,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0249550342559814,
+      "learning_rate": 0.00046109999999999993,
+      "loss": 2.5392,
       "step": 2950
     },
     {
       "epoch": 4.7216000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 15.6370210647583,
+      "learning_rate": 0.00046094999999999996,
+      "loss": 2.5412,
       "step": 2951
     },
     {
       "epoch": 4.7232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 12.782587051391602,
+      "learning_rate": 0.0004608,
+      "loss": 2.6349,
       "step": 2952
     },
     {
       "epoch": 4.7248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.132637023925781,
+      "learning_rate": 0.00046065,
+      "loss": 2.2466,
       "step": 2953
     },
     {
       "epoch": 4.7264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 15.21384048461914,
+      "learning_rate": 0.0004604999999999999,
+      "loss": 2.4299,
       "step": 2954
     },
     {
       "epoch": 4.728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.304612159729004,
+      "learning_rate": 0.00046034999999999994,
+      "loss": 1.9851,
       "step": 2955
     },
     {
       "epoch": 4.7296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.457035541534424,
+      "learning_rate": 0.00046019999999999996,
+      "loss": 1.9655,
       "step": 2956
     },
     {
       "epoch": 4.7312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.9795613288879395,
+      "learning_rate": 0.00046005,
+      "loss": 1.9755,
       "step": 2957
     },
     {
       "epoch": 4.7328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.6748929023742676,
+      "learning_rate": 0.00045989999999999996,
+      "loss": 1.9936,
       "step": 2958
     },
     {
       "epoch": 4.7344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5004043579101562,
+      "learning_rate": 0.0004597499999999999,
+      "loss": 2.254,
       "step": 2959
     },
     {
       "epoch": 4.736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3329801559448242,
+      "learning_rate": 0.00045959999999999995,
+      "loss": 1.7511,
       "step": 2960
     },
     {
       "epoch": 4.7376000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.246700286865234,
+      "learning_rate": 0.00045945,
+      "loss": 1.6901,
       "step": 2961
     },
     {
       "epoch": 4.7392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.481882095336914,
+      "learning_rate": 0.00045929999999999994,
+      "loss": 2.0359,
       "step": 2962
     },
     {
       "epoch": 4.7408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 12.636874198913574,
+      "learning_rate": 0.00045914999999999997,
+      "loss": 2.1203,
       "step": 2963
     },
     {
       "epoch": 4.7424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.530386924743652,
+      "learning_rate": 0.000459,
+      "loss": 1.78,
       "step": 2964
     },
     {
       "epoch": 4.744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.78328275680542,
+      "learning_rate": 0.00045884999999999996,
+      "loss": 1.8293,
       "step": 2965
     },
     {
       "epoch": 4.7456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4835050106048584,
+      "learning_rate": 0.00045869999999999993,
+      "loss": 1.9996,
       "step": 2966
     },
     {
       "epoch": 4.7472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7223535776138306,
+      "learning_rate": 0.00045854999999999995,
+      "loss": 2.1396,
       "step": 2967
     },
     {
       "epoch": 4.7488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6066192388534546,
+      "learning_rate": 0.0004584,
+      "loss": 1.7808,
       "step": 2968
     },
     {
       "epoch": 4.7504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.09505558013916,
+      "learning_rate": 0.00045825,
+      "loss": 1.6402,
       "step": 2969
     },
     {
       "epoch": 4.752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1714239120483398,
+      "learning_rate": 0.0004580999999999999,
+      "loss": 1.8108,
       "step": 2970
     },
     {
       "epoch": 4.7536000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.506039619445801,
+      "learning_rate": 0.00045794999999999994,
+      "loss": 1.9285,
       "step": 2971
     },
     {
       "epoch": 4.7552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0846710205078125,
+      "learning_rate": 0.00045779999999999996,
+      "loss": 1.6564,
       "step": 2972
     },
     {
       "epoch": 4.7568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5491540431976318,
+      "learning_rate": 0.00045765,
+      "loss": 1.6533,
       "step": 2973
     },
     {
       "epoch": 4.7584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.191540241241455,
+      "learning_rate": 0.00045749999999999995,
+      "loss": 1.6933,
       "step": 2974
     },
     {
       "epoch": 4.76,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.992797374725342,
+      "learning_rate": 0.00045735,
+      "loss": 1.886,
       "step": 2975
     },
     {
       "epoch": 4.7616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2760454416275024,
+      "learning_rate": 0.00045719999999999995,
+      "loss": 1.9759,
       "step": 2976
     },
     {
       "epoch": 4.7632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1154367923736572,
+      "learning_rate": 0.00045704999999999997,
+      "loss": 1.8352,
       "step": 2977
     },
     {
       "epoch": 4.7648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2374048233032227,
+      "learning_rate": 0.00045689999999999994,
+      "loss": 2.0876,
       "step": 2978
     },
     {
       "epoch": 4.7664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.916267991065979,
+      "learning_rate": 0.00045674999999999996,
+      "loss": 1.8537,
       "step": 2979
     },
     {
       "epoch": 4.768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9522725343704224,
+      "learning_rate": 0.0004566,
+      "loss": 1.8254,
       "step": 2980
     },
     {
       "epoch": 4.7696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4546539783477783,
+      "learning_rate": 0.00045645,
+      "loss": 2.0879,
       "step": 2981
     },
     {
       "epoch": 4.7712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.203273296356201,
+      "learning_rate": 0.0004562999999999999,
+      "loss": 2.0429,
       "step": 2982
     },
     {
       "epoch": 4.7728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.57181978225708,
+      "learning_rate": 0.00045614999999999995,
+      "loss": 2.3272,
       "step": 2983
     },
     {
       "epoch": 4.7744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5816924571990967,
+      "learning_rate": 0.00045599999999999997,
+      "loss": 1.9874,
       "step": 2984
     },
     {
       "epoch": 4.776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3904832601547241,
+      "learning_rate": 0.00045585,
+      "loss": 1.8478,
       "step": 2985
     },
     {
       "epoch": 4.7776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9760186672210693,
+      "learning_rate": 0.0004556999999999999,
+      "loss": 2.0731,
       "step": 2986
     },
     {
       "epoch": 4.7792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3998148441314697,
+      "learning_rate": 0.00045554999999999993,
+      "loss": 2.1853,
       "step": 2987
     },
     {
       "epoch": 4.7808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3389252424240112,
+      "learning_rate": 0.00045539999999999996,
+      "loss": 2.0391,
       "step": 2988
     },
     {
       "epoch": 4.7824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.23905086517334,
+      "learning_rate": 0.00045525,
+      "loss": 2.2752,
       "step": 2989
     },
     {
       "epoch": 4.784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6538360118865967,
+      "learning_rate": 0.00045509999999999995,
+      "loss": 1.9852,
       "step": 2990
     },
     {
       "epoch": 4.7856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.154603004455566,
+      "learning_rate": 0.00045494999999999997,
+      "loss": 1.9189,
       "step": 2991
     },
     {
       "epoch": 4.7872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.789768695831299,
+      "learning_rate": 0.00045479999999999994,
+      "loss": 2.0853,
       "step": 2992
     },
     {
       "epoch": 4.7888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.0123066902160645,
+      "learning_rate": 0.00045464999999999997,
+      "loss": 2.1091,
       "step": 2993
     },
     {
       "epoch": 4.7904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.259772777557373,
+      "learning_rate": 0.00045449999999999993,
+      "loss": 1.8027,
       "step": 2994
     },
     {
       "epoch": 4.792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9710586071014404,
+      "learning_rate": 0.00045434999999999996,
+      "loss": 2.2043,
       "step": 2995
     },
     {
       "epoch": 4.7936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4823061227798462,
+      "learning_rate": 0.0004542,
+      "loss": 1.6574,
       "step": 2996
     },
     {
       "epoch": 4.7952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.3044240474700928,
+      "learning_rate": 0.00045405,
+      "loss": 2.3636,
       "step": 2997
     },
     {
       "epoch": 4.7968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.694255828857422,
+      "learning_rate": 0.0004538999999999999,
+      "loss": 2.4637,
       "step": 2998
     },
     {
       "epoch": 4.7984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.3031346797943115,
+      "learning_rate": 0.00045374999999999994,
+      "loss": 2.1207,
       "step": 2999
     },
     {
       "epoch": 4.8,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5991824865341187,
+      "learning_rate": 0.00045359999999999997,
+      "loss": 2.3755,
       "step": 3000
     },
     {
       "epoch": 4.8,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 160.2639,
-      "eval_samples_per_second": 19.568,
+      "eval_cer": 0.4432534778969187,
+      "eval_loss": 2.37933349609375,
+      "eval_runtime": 160.2371,
+      "eval_samples_per_second": 19.571,
       "eval_steps_per_second": 1.223,
-      "eval_wer": 1.0,
+      "eval_wer": 0.7270127447932857,
       "step": 3000
     },
     {
       "epoch": 4.8016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 24.620410919189453,
+      "learning_rate": 0.00045345,
+      "loss": 3.1706,
       "step": 3001
     },
     {
       "epoch": 4.8032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.307674407958984,
+      "learning_rate": 0.0004532999999999999,
+      "loss": 2.1999,
       "step": 3002
     },
     {
       "epoch": 4.8048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 23.469606399536133,
+      "learning_rate": 0.00045314999999999993,
+      "loss": 3.5088,
       "step": 3003
     },
     {
       "epoch": 4.8064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.592309832572937,
+      "learning_rate": 0.00045299999999999995,
+      "loss": 2.1195,
       "step": 3004
     },
     {
       "epoch": 4.808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.467976093292236,
+      "learning_rate": 0.00045285,
+      "loss": 2.3242,
       "step": 3005
     },
     {
       "epoch": 4.8096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7775919437408447,
+      "learning_rate": 0.00045269999999999994,
+      "loss": 1.9178,
       "step": 3006
     },
     {
       "epoch": 4.8112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.592406272888184,
+      "learning_rate": 0.00045254999999999997,
+      "loss": 2.5356,
       "step": 3007
     },
     {
       "epoch": 4.8128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.714765548706055,
+      "learning_rate": 0.00045239999999999994,
+      "loss": 1.8963,
       "step": 3008
     },
     {
       "epoch": 4.8144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2107770442962646,
+      "learning_rate": 0.00045224999999999996,
+      "loss": 1.5007,
       "step": 3009
     },
     {
       "epoch": 4.816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.1892650127410889,
+      "learning_rate": 0.00045209999999999993,
+      "loss": 1.7101,
       "step": 3010
     },
     {
       "epoch": 4.8176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.4028902053833,
+      "learning_rate": 0.00045194999999999995,
+      "loss": 2.0486,
       "step": 3011
     },
     {
       "epoch": 4.8192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.556225299835205,
+      "learning_rate": 0.0004518,
+      "loss": 1.8077,
       "step": 3012
     },
     {
       "epoch": 4.8208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.0722336769104,
+      "learning_rate": 0.00045165,
+      "loss": 1.9737,
       "step": 3013
     },
     {
       "epoch": 4.8224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.667191505432129,
+      "learning_rate": 0.0004514999999999999,
+      "loss": 1.7824,
       "step": 3014
     },
     {
       "epoch": 4.824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2074222564697266,
+      "learning_rate": 0.00045134999999999994,
+      "loss": 1.8937,
       "step": 3015
     },
     {
       "epoch": 4.8256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.051270484924316,
+      "learning_rate": 0.00045119999999999996,
+      "loss": 1.846,
       "step": 3016
     },
     {
       "epoch": 4.8272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.784696340560913,
+      "learning_rate": 0.00045105,
+      "loss": 1.7694,
       "step": 3017
     },
     {
       "epoch": 4.8288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.480818748474121,
+      "learning_rate": 0.0004508999999999999,
+      "loss": 1.9753,
       "step": 3018
     },
     {
       "epoch": 4.8304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1542861461639404,
+      "learning_rate": 0.0004507499999999999,
+      "loss": 1.6415,
       "step": 3019
     },
     {
       "epoch": 4.832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.286095142364502,
+      "learning_rate": 0.00045059999999999995,
+      "loss": 2.0295,
       "step": 3020
     },
     {
       "epoch": 4.8336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7635704278945923,
+      "learning_rate": 0.00045044999999999997,
+      "loss": 1.6449,
       "step": 3021
     },
     {
       "epoch": 4.8352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3237788677215576,
+      "learning_rate": 0.00045029999999999994,
+      "loss": 2.4472,
       "step": 3022
     },
     {
       "epoch": 4.8368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.521028995513916,
+      "learning_rate": 0.00045014999999999996,
+      "loss": 1.9126,
       "step": 3023
     },
     {
       "epoch": 4.8384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.178478717803955,
+      "learning_rate": 0.00045,
+      "loss": 2.2155,
       "step": 3024
     },
     {
       "epoch": 4.84,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7699393033981323,
+      "learning_rate": 0.00044984999999999996,
+      "loss": 1.9071,
       "step": 3025
     },
     {
       "epoch": 4.8416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4859039783477783,
+      "learning_rate": 0.0004497,
+      "loss": 1.7341,
       "step": 3026
     },
     {
       "epoch": 4.8431999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2301384210586548,
+      "learning_rate": 0.00044954999999999995,
+      "loss": 1.6586,
       "step": 3027
     },
     {
       "epoch": 4.8448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.274512529373169,
+      "learning_rate": 0.0004494,
+      "loss": 2.1292,
       "step": 3028
     },
     {
       "epoch": 4.8464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0131479501724243,
+      "learning_rate": 0.00044925,
+      "loss": 1.7007,
       "step": 3029
     },
     {
       "epoch": 4.848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.0874526500701904,
+      "learning_rate": 0.0004491,
+      "loss": 1.692,
       "step": 3030
     },
     {
       "epoch": 4.8496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2531495094299316,
+      "learning_rate": 0.00044894999999999994,
+      "loss": 1.7062,
       "step": 3031
     },
     {
       "epoch": 4.8512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9190071821212769,
+      "learning_rate": 0.00044879999999999996,
+      "loss": 1.8297,
       "step": 3032
     },
     {
       "epoch": 4.8528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5107418298721313,
+      "learning_rate": 0.00044865,
+      "loss": 1.8974,
       "step": 3033
     },
     {
       "epoch": 4.8544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.797076940536499,
+      "learning_rate": 0.0004485,
+      "loss": 2.1044,
       "step": 3034
     },
     {
       "epoch": 4.856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9107513427734375,
+      "learning_rate": 0.0004483499999999999,
+      "loss": 1.901,
       "step": 3035
     },
     {
       "epoch": 4.8576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.097909927368164,
+      "learning_rate": 0.00044819999999999994,
+      "loss": 1.984,
       "step": 3036
     },
     {
       "epoch": 4.8591999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7260329723358154,
+      "learning_rate": 0.00044804999999999997,
+      "loss": 1.6391,
       "step": 3037
     },
     {
       "epoch": 4.8608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.3167763948440552,
+      "learning_rate": 0.0004479,
+      "loss": 1.8456,
       "step": 3038
     },
     {
       "epoch": 4.8624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.121819019317627,
+      "learning_rate": 0.00044774999999999996,
+      "loss": 2.1833,
       "step": 3039
     },
     {
       "epoch": 4.864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3631556034088135,
+      "learning_rate": 0.0004476,
+      "loss": 1.5566,
       "step": 3040
     },
     {
       "epoch": 4.8656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2354460954666138,
+      "learning_rate": 0.00044744999999999995,
+      "loss": 1.9648,
       "step": 3041
     },
     {
       "epoch": 4.8672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4966917037963867,
+      "learning_rate": 0.0004473,
+      "loss": 1.7334,
       "step": 3042
     },
     {
       "epoch": 4.8688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9379713535308838,
+      "learning_rate": 0.00044714999999999995,
+      "loss": 2.5152,
       "step": 3043
     },
     {
       "epoch": 4.8704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.353835105895996,
+      "learning_rate": 0.00044699999999999997,
+      "loss": 1.9858,
       "step": 3044
     },
     {
       "epoch": 4.872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.308361291885376,
+      "learning_rate": 0.00044685,
+      "loss": 1.7178,
       "step": 3045
     },
     {
       "epoch": 4.8736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9239517450332642,
+      "learning_rate": 0.0004467,
+      "loss": 2.0712,
       "step": 3046
     },
     {
       "epoch": 4.8751999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6877199411392212,
+      "learning_rate": 0.00044654999999999993,
+      "loss": 1.9809,
       "step": 3047
     },
     {
       "epoch": 4.8768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.460870742797852,
+      "learning_rate": 0.00044639999999999995,
+      "loss": 2.3111,
       "step": 3048
     },
     {
       "epoch": 4.8784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2702107429504395,
+      "learning_rate": 0.00044625,
+      "loss": 2.2167,
       "step": 3049
     },
     {
       "epoch": 4.88,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00044625,
+      "loss": 2.7315,
       "step": 3050
     },
     {
       "epoch": 4.8816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 16.57255744934082,
+      "learning_rate": 0.0004461,
+      "loss": 2.9102,
       "step": 3051
     },
     {
       "epoch": 4.8832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.352267265319824,
+      "learning_rate": 0.0004459499999999999,
+      "loss": 2.5403,
       "step": 3052
     },
     {
       "epoch": 4.8848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.948455810546875,
+      "learning_rate": 0.00044579999999999994,
+      "loss": 2.0783,
       "step": 3053
     },
     {
       "epoch": 4.8864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.270328998565674,
+      "learning_rate": 0.00044564999999999996,
+      "loss": 2.7684,
       "step": 3054
     },
     {
       "epoch": 4.888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1044552326202393,
+      "learning_rate": 0.0004455,
+      "loss": 2.26,
       "step": 3055
     },
     {
       "epoch": 4.8896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.126490592956543,
+      "learning_rate": 0.00044534999999999996,
+      "loss": 1.9853,
       "step": 3056
     },
     {
       "epoch": 4.8911999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.923595905303955,
+      "learning_rate": 0.0004452,
+      "loss": 2.5213,
       "step": 3057
     },
     {
       "epoch": 4.8928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.534145832061768,
+      "learning_rate": 0.00044504999999999995,
+      "loss": 1.9294,
       "step": 3058
     },
     {
       "epoch": 4.8944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.250170707702637,
+      "learning_rate": 0.00044489999999999997,
+      "loss": 2.1167,
       "step": 3059
     },
     {
       "epoch": 4.896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8796231746673584,
+      "learning_rate": 0.00044474999999999994,
+      "loss": 2.0915,
       "step": 3060
     },
     {
       "epoch": 4.8976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8455438613891602,
+      "learning_rate": 0.00044459999999999996,
+      "loss": 2.5508,
       "step": 3061
     },
     {
       "epoch": 4.8992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.386833190917969,
+      "learning_rate": 0.00044445,
+      "loss": 1.7572,
       "step": 3062
     },
     {
       "epoch": 4.9008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.531754016876221,
+      "learning_rate": 0.0004443,
+      "loss": 1.9147,
       "step": 3063
     },
     {
       "epoch": 4.9024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6494725942611694,
+      "learning_rate": 0.0004441499999999999,
+      "loss": 1.7609,
       "step": 3064
     },
     {
       "epoch": 4.904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8058489561080933,
+      "learning_rate": 0.00044399999999999995,
+      "loss": 1.8494,
       "step": 3065
     },
     {
       "epoch": 4.9056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1291446685791016,
+      "learning_rate": 0.00044385,
+      "loss": 1.6648,
       "step": 3066
     },
     {
       "epoch": 4.9072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5931782722473145,
+      "learning_rate": 0.0004437,
+      "loss": 1.8376,
       "step": 3067
     },
     {
       "epoch": 4.9088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7532620429992676,
+      "learning_rate": 0.0004435499999999999,
+      "loss": 1.7616,
       "step": 3068
     },
     {
       "epoch": 4.9104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9374960660934448,
+      "learning_rate": 0.00044339999999999994,
+      "loss": 1.8737,
       "step": 3069
     },
     {
       "epoch": 4.912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7475881576538086,
+      "learning_rate": 0.00044324999999999996,
+      "loss": 1.9106,
       "step": 3070
     },
     {
       "epoch": 4.9136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.425243854522705,
+      "learning_rate": 0.0004431,
+      "loss": 1.5245,
       "step": 3071
     },
     {
       "epoch": 4.9152000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2945805788040161,
+      "learning_rate": 0.00044294999999999995,
+      "loss": 1.8072,
       "step": 3072
     },
     {
       "epoch": 4.9168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6314753293991089,
+      "learning_rate": 0.0004428,
+      "loss": 2.1144,
       "step": 3073
     },
     {
       "epoch": 4.9184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.375238060951233,
+      "learning_rate": 0.00044264999999999994,
+      "loss": 2.2158,
       "step": 3074
     },
     {
       "epoch": 4.92,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8653980493545532,
+      "learning_rate": 0.00044249999999999997,
+      "loss": 1.9081,
       "step": 3075
     },
     {
       "epoch": 4.9216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.507571697235107,
+      "learning_rate": 0.00044234999999999994,
+      "loss": 1.8373,
       "step": 3076
     },
     {
       "epoch": 4.9232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9098926782608032,
+      "learning_rate": 0.00044219999999999996,
+      "loss": 1.7228,
       "step": 3077
     },
     {
       "epoch": 4.9248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.42020845413208,
+      "learning_rate": 0.00044205,
+      "loss": 1.6555,
       "step": 3078
     },
     {
       "epoch": 4.9264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.94321346282959,
+      "learning_rate": 0.0004419,
+      "loss": 1.7539,
       "step": 3079
     },
     {
       "epoch": 4.928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5416208505630493,
+      "learning_rate": 0.0004417499999999999,
+      "loss": 1.4654,
       "step": 3080
     },
     {
       "epoch": 4.9296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7197470664978027,
+      "learning_rate": 0.00044159999999999995,
+      "loss": 2.085,
       "step": 3081
     },
     {
       "epoch": 4.9312000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6691932678222656,
+      "learning_rate": 0.00044144999999999997,
+      "loss": 2.0641,
       "step": 3082
     },
     {
       "epoch": 4.9328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6235414743423462,
+      "learning_rate": 0.0004413,
+      "loss": 2.285,
       "step": 3083
     },
     {
       "epoch": 4.9344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4483048915863037,
+      "learning_rate": 0.00044114999999999996,
+      "loss": 2.0029,
       "step": 3084
     },
     {
       "epoch": 4.936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2496280670166016,
+      "learning_rate": 0.00044099999999999993,
+      "loss": 1.6248,
       "step": 3085
     },
     {
       "epoch": 4.9376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6654266119003296,
+      "learning_rate": 0.00044084999999999996,
+      "loss": 2.0614,
       "step": 3086
     },
     {
       "epoch": 4.9392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0901358127593994,
+      "learning_rate": 0.0004407,
+      "loss": 2.2341,
       "step": 3087
     },
     {
       "epoch": 4.9408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6806238889694214,
+      "learning_rate": 0.00044054999999999995,
+      "loss": 1.7547,
       "step": 3088
     },
     {
       "epoch": 4.9424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5384247303009033,
+      "learning_rate": 0.00044039999999999997,
+      "loss": 1.7914,
       "step": 3089
     },
     {
       "epoch": 4.944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.296206951141357,
+      "learning_rate": 0.00044025,
+      "loss": 1.9357,
       "step": 3090
     },
     {
       "epoch": 4.9456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6634535789489746,
+      "learning_rate": 0.00044009999999999996,
+      "loss": 1.9188,
       "step": 3091
     },
     {
       "epoch": 4.9472000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.649627923965454,
+      "learning_rate": 0.00043994999999999993,
+      "loss": 1.9103,
       "step": 3092
     },
     {
       "epoch": 4.9488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5803511142730713,
+      "learning_rate": 0.00043979999999999996,
+      "loss": 1.8764,
       "step": 3093
     },
     {
       "epoch": 4.9504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5392707586288452,
+      "learning_rate": 0.00043965,
+      "loss": 2.0883,
       "step": 3094
     },
     {
       "epoch": 4.952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.35639488697052,
+      "learning_rate": 0.0004395,
+      "loss": 2.2488,
       "step": 3095
     },
     {
       "epoch": 4.9536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5602680444717407,
+      "learning_rate": 0.0004393499999999999,
+      "loss": 1.6585,
       "step": 3096
     },
     {
       "epoch": 4.9552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.285512924194336,
+      "learning_rate": 0.00043919999999999994,
+      "loss": 2.3513,
       "step": 3097
     },
     {
       "epoch": 4.9568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6870580911636353,
+      "learning_rate": 0.00043904999999999997,
+      "loss": 2.8381,
       "step": 3098
     },
     {
       "epoch": 4.9584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5827438831329346,
+      "learning_rate": 0.0004389,
+      "loss": 2.2782,
       "step": 3099
     },
     {
       "epoch": 4.96,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 3100
-    },
-    {
-      "epoch": 4.96,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 156.866,
-      "eval_samples_per_second": 19.992,
-      "eval_steps_per_second": 1.249,
-      "eval_wer": 1.0,
+      "learning_rate": 0.0004389,
+      "loss": 1.7011,
       "step": 3100
     },
     {
       "epoch": 4.9616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.932778358459473,
+      "learning_rate": 0.00043874999999999996,
+      "loss": 2.8354,
       "step": 3101
     },
     {
       "epoch": 4.9632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.474605560302734,
+      "learning_rate": 0.00043859999999999993,
+      "loss": 3.0243,
       "step": 3102
     },
     {
       "epoch": 4.9648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.513943195343018,
+      "learning_rate": 0.00043844999999999995,
+      "loss": 2.1145,
       "step": 3103
     },
     {
       "epoch": 4.9664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.522618293762207,
+      "learning_rate": 0.0004383,
+      "loss": 2.5017,
       "step": 3104
     },
     {
       "epoch": 4.968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.3348541259765625,
+      "learning_rate": 0.00043814999999999994,
+      "loss": 1.8663,
       "step": 3105
     },
     {
       "epoch": 4.9696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.547016143798828,
+      "learning_rate": 0.00043799999999999997,
+      "loss": 1.9593,
       "step": 3106
     },
     {
       "epoch": 4.9712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8356808423995972,
+      "learning_rate": 0.00043785,
+      "loss": 1.6838,
       "step": 3107
     },
     {
       "epoch": 4.9728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.7317214012145996,
+      "learning_rate": 0.00043769999999999996,
+      "loss": 1.8651,
       "step": 3108
     },
     {
       "epoch": 4.9744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4681026935577393,
+      "learning_rate": 0.00043754999999999993,
+      "loss": 2.0261,
       "step": 3109
     },
     {
       "epoch": 4.976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7375612258911133,
+      "learning_rate": 0.00043739999999999995,
+      "loss": 1.9152,
       "step": 3110
     },
     {
       "epoch": 4.9776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.383476972579956,
+      "learning_rate": 0.00043725,
+      "loss": 2.047,
       "step": 3111
     },
     {
       "epoch": 4.9792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.8656511306762695,
+      "learning_rate": 0.0004371,
+      "loss": 2.43,
       "step": 3112
     },
     {
       "epoch": 4.9808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.090378761291504,
+      "learning_rate": 0.0004369499999999999,
+      "loss": 2.1424,
       "step": 3113
     },
     {
       "epoch": 4.9824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.096062421798706,
+      "learning_rate": 0.00043679999999999994,
+      "loss": 1.7151,
       "step": 3114
     },
     {
       "epoch": 4.984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4025139808654785,
+      "learning_rate": 0.00043664999999999996,
+      "loss": 1.809,
       "step": 3115
     },
     {
       "epoch": 4.9856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3236799240112305,
+      "learning_rate": 0.0004365,
+      "loss": 2.0313,
       "step": 3116
     },
     {
       "epoch": 4.9872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.579554557800293,
+      "learning_rate": 0.00043634999999999995,
+      "loss": 1.7257,
       "step": 3117
     },
     {
       "epoch": 4.9888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1674931049346924,
+      "learning_rate": 0.0004361999999999999,
+      "loss": 2.0792,
       "step": 3118
     },
     {
       "epoch": 4.9904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9446090459823608,
+      "learning_rate": 0.00043604999999999995,
+      "loss": 1.8795,
       "step": 3119
     },
     {
       "epoch": 4.992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.445674419403076,
+      "learning_rate": 0.00043589999999999997,
+      "loss": 2.0259,
       "step": 3120
     },
     {
       "epoch": 4.9936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.725457668304443,
+      "learning_rate": 0.00043574999999999994,
+      "loss": 2.0193,
       "step": 3121
     },
     {
       "epoch": 4.9952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.952589988708496,
+      "learning_rate": 0.00043559999999999996,
+      "loss": 2.2579,
       "step": 3122
     },
     {
       "epoch": 4.9968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.328230142593384,
+      "learning_rate": 0.00043545,
+      "loss": 2.8075,
       "step": 3123
     },
     {
       "epoch": 4.9984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.565820574760437,
+      "learning_rate": 0.00043529999999999996,
+      "loss": 2.2939,
       "step": 3124
     },
     {
       "epoch": 5.0,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.119185209274292,
+      "learning_rate": 0.0004351499999999999,
+      "loss": 2.4498,
       "step": 3125
     },
     {
       "epoch": 5.0016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.970066547393799,
+      "learning_rate": 0.00043499999999999995,
+      "loss": 2.4477,
       "step": 3126
     },
     {
       "epoch": 5.0032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.783778190612793,
+      "learning_rate": 0.00043484999999999997,
+      "loss": 2.3359,
       "step": 3127
     },
     {
       "epoch": 5.0048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 19.11392593383789,
+      "learning_rate": 0.0004347,
+      "loss": 2.7559,
       "step": 3128
     },
     {
       "epoch": 5.0064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7798497676849365,
+      "learning_rate": 0.0004345499999999999,
+      "loss": 2.0971,
       "step": 3129
     },
     {
       "epoch": 5.008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.793581962585449,
+      "learning_rate": 0.00043439999999999993,
+      "loss": 2.6192,
       "step": 3130
     },
     {
       "epoch": 5.0096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.100231170654297,
+      "learning_rate": 0.00043424999999999996,
+      "loss": 1.9119,
       "step": 3131
     },
     {
       "epoch": 5.0112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.948700904846191,
+      "learning_rate": 0.0004341,
+      "loss": 2.3903,
       "step": 3132
     },
     {
       "epoch": 5.0128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.955105781555176,
+      "learning_rate": 0.00043394999999999995,
+      "loss": 2.0834,
       "step": 3133
     },
     {
       "epoch": 5.0144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.116028070449829,
+      "learning_rate": 0.0004338,
+      "loss": 1.9087,
       "step": 3134
     },
     {
       "epoch": 5.016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.022517442703247,
+      "learning_rate": 0.00043364999999999994,
+      "loss": 1.8392,
       "step": 3135
     },
     {
       "epoch": 5.0176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.67773962020874,
+      "learning_rate": 0.00043349999999999997,
+      "loss": 1.9662,
       "step": 3136
     },
     {
       "epoch": 5.0192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.792007446289062,
+      "learning_rate": 0.00043334999999999994,
+      "loss": 2.2113,
       "step": 3137
     },
     {
       "epoch": 5.0208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.263298034667969,
+      "learning_rate": 0.00043319999999999996,
+      "loss": 1.8671,
       "step": 3138
     },
     {
       "epoch": 5.0224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 11.039977073669434,
+      "learning_rate": 0.00043305,
+      "loss": 2.3769,
       "step": 3139
     },
     {
       "epoch": 5.024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.120280265808105,
+      "learning_rate": 0.0004329,
+      "loss": 2.0752,
       "step": 3140
     },
     {
       "epoch": 5.0256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.3145546913146973,
+      "learning_rate": 0.0004327499999999999,
+      "loss": 1.9827,
       "step": 3141
     },
     {
       "epoch": 5.0272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8486971855163574,
+      "learning_rate": 0.00043259999999999994,
+      "loss": 1.7318,
       "step": 3142
     },
     {
       "epoch": 5.0288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 11.292924880981445,
+      "learning_rate": 0.00043244999999999997,
+      "loss": 2.1877,
       "step": 3143
     },
     {
       "epoch": 5.0304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.601168632507324,
+      "learning_rate": 0.0004323,
+      "loss": 1.7993,
       "step": 3144
     },
     {
       "epoch": 5.032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0242254734039307,
+      "learning_rate": 0.0004321499999999999,
+      "loss": 1.6663,
       "step": 3145
     },
     {
       "epoch": 5.0336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0984086990356445,
+      "learning_rate": 0.00043199999999999993,
+      "loss": 1.9235,
       "step": 3146
     },
     {
       "epoch": 5.0352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6991688013076782,
+      "learning_rate": 0.00043184999999999995,
+      "loss": 1.6364,
       "step": 3147
     },
     {
       "epoch": 5.0368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7893872261047363,
+      "learning_rate": 0.0004317,
+      "loss": 2.1728,
       "step": 3148
     },
     {
       "epoch": 5.0384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.034099102020264,
+      "learning_rate": 0.00043154999999999995,
+      "loss": 2.162,
       "step": 3149
     },
     {
       "epoch": 5.04,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.309537410736084,
+      "learning_rate": 0.00043139999999999997,
+      "loss": 1.7779,
       "step": 3150
     },
     {
       "epoch": 5.0416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.693896532058716,
+      "learning_rate": 0.00043124999999999994,
+      "loss": 1.6807,
       "step": 3151
     },
     {
       "epoch": 5.0432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.089795112609863,
+      "learning_rate": 0.00043109999999999996,
+      "loss": 1.9001,
       "step": 3152
     },
     {
       "epoch": 5.0448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.010385513305664,
+      "learning_rate": 0.00043095,
+      "loss": 1.5703,
       "step": 3153
     },
     {
       "epoch": 5.0464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8490614891052246,
+      "learning_rate": 0.00043079999999999995,
+      "loss": 1.9197,
       "step": 3154
     },
     {
       "epoch": 5.048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.640694499015808,
+      "learning_rate": 0.00043065,
+      "loss": 1.6976,
       "step": 3155
     },
     {
       "epoch": 5.0496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.294865846633911,
+      "learning_rate": 0.0004305,
+      "loss": 1.8106,
       "step": 3156
     },
     {
       "epoch": 5.0512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1499030590057373,
+      "learning_rate": 0.00043034999999999997,
+      "loss": 2.1928,
       "step": 3157
     },
     {
       "epoch": 5.0528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.309077739715576,
+      "learning_rate": 0.00043019999999999994,
+      "loss": 1.7319,
       "step": 3158
     },
     {
       "epoch": 5.0544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.6867856979370117,
+      "learning_rate": 0.00043004999999999996,
+      "loss": 2.0171,
       "step": 3159
     },
     {
       "epoch": 5.056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8424367904663086,
+      "learning_rate": 0.0004299,
+      "loss": 2.0746,
       "step": 3160
     },
     {
       "epoch": 5.0576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5826704502105713,
+      "learning_rate": 0.00042975,
+      "loss": 2.0328,
       "step": 3161
     },
     {
       "epoch": 5.0592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.853431224822998,
+      "learning_rate": 0.0004295999999999999,
+      "loss": 1.9964,
       "step": 3162
     },
     {
       "epoch": 5.0608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6717617511749268,
+      "learning_rate": 0.00042944999999999995,
+      "loss": 2.0824,
       "step": 3163
     },
     {
       "epoch": 5.0624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.649266004562378,
+      "learning_rate": 0.00042929999999999997,
+      "loss": 2.3392,
       "step": 3164
     },
     {
       "epoch": 5.064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8702266216278076,
+      "learning_rate": 0.00042915,
+      "loss": 2.1132,
       "step": 3165
     },
     {
       "epoch": 5.0656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.922991752624512,
+      "learning_rate": 0.00042899999999999997,
+      "loss": 2.4097,
       "step": 3166
     },
     {
       "epoch": 5.0672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.757828712463379,
+      "learning_rate": 0.00042884999999999993,
+      "loss": 2.3162,
       "step": 3167
     },
     {
       "epoch": 5.0688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.166485071182251,
+      "learning_rate": 0.00042869999999999996,
+      "loss": 2.0998,
       "step": 3168
     },
     {
       "epoch": 5.0704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8233377933502197,
+      "learning_rate": 0.00042855,
+      "loss": 2.1253,
       "step": 3169
     },
     {
       "epoch": 5.072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8624346256256104,
+      "learning_rate": 0.00042839999999999995,
+      "loss": 2.0939,
       "step": 3170
     },
     {
       "epoch": 5.0736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.785826683044434,
+      "learning_rate": 0.00042825,
+      "loss": 2.0271,
       "step": 3171
     },
     {
       "epoch": 5.0752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.498620510101318,
+      "learning_rate": 0.0004281,
+      "loss": 2.0795,
       "step": 3172
     },
     {
       "epoch": 5.0768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1600935459136963,
+      "learning_rate": 0.00042794999999999997,
+      "loss": 2.3353,
       "step": 3173
     },
     {
       "epoch": 5.0784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.206274032592773,
+      "learning_rate": 0.00042779999999999994,
+      "loss": 2.7881,
       "step": 3174
     },
     {
       "epoch": 5.08,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.14153790473938,
+      "learning_rate": 0.00042764999999999996,
+      "loss": 2.886,
       "step": 3175
     },
     {
       "epoch": 5.0816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.464966773986816,
+      "learning_rate": 0.0004275,
+      "loss": 2.7726,
       "step": 3176
     },
     {
       "epoch": 5.0832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2620949745178223,
+      "learning_rate": 0.00042735,
+      "loss": 2.3795,
       "step": 3177
     },
     {
       "epoch": 5.0848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8024942874908447,
+      "learning_rate": 0.0004271999999999999,
+      "loss": 2.4685,
       "step": 3178
     },
     {
       "epoch": 5.0864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.8808012008667,
+      "learning_rate": 0.00042704999999999994,
+      "loss": 3.0535,
       "step": 3179
     },
     {
       "epoch": 5.088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9684244394302368,
+      "learning_rate": 0.00042689999999999997,
+      "loss": 2.4331,
       "step": 3180
     },
     {
       "epoch": 5.0896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.007668495178223,
+      "learning_rate": 0.00042675,
+      "loss": 2.2903,
       "step": 3181
     },
     {
       "epoch": 5.0912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.786413192749023,
+      "learning_rate": 0.00042659999999999996,
+      "loss": 2.0783,
       "step": 3182
     },
     {
       "epoch": 5.0928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8741869926452637,
+      "learning_rate": 0.00042645,
+      "loss": 2.7336,
       "step": 3183
     },
     {
       "epoch": 5.0944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.79109001159668,
+      "learning_rate": 0.00042629999999999995,
+      "loss": 2.399,
       "step": 3184
     },
     {
       "epoch": 5.096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3625593185424805,
+      "learning_rate": 0.00042615,
+      "loss": 1.9723,
       "step": 3185
     },
     {
       "epoch": 5.0976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4920530319213867,
+      "learning_rate": 0.00042599999999999995,
+      "loss": 2.0308,
       "step": 3186
     },
     {
       "epoch": 5.0992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7185099124908447,
+      "learning_rate": 0.00042584999999999997,
+      "loss": 2.0599,
       "step": 3187
     },
     {
       "epoch": 5.1008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5361008644104004,
+      "learning_rate": 0.0004257,
+      "loss": 1.9269,
       "step": 3188
     },
     {
       "epoch": 5.1024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.932170033454895,
+      "learning_rate": 0.00042555,
+      "loss": 2.0408,
       "step": 3189
     },
     {
       "epoch": 5.104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8861477375030518,
+      "learning_rate": 0.00042539999999999993,
+      "loss": 2.1108,
       "step": 3190
     },
     {
       "epoch": 5.1056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9711568355560303,
+      "learning_rate": 0.00042524999999999996,
+      "loss": 1.92,
       "step": 3191
     },
     {
       "epoch": 5.1072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2188079357147217,
+      "learning_rate": 0.0004251,
+      "loss": 2.0846,
       "step": 3192
     },
     {
       "epoch": 5.1088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8576302528381348,
+      "learning_rate": 0.00042495,
+      "loss": 2.1029,
       "step": 3193
     },
     {
       "epoch": 5.1104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.3662407398223877,
+      "learning_rate": 0.0004247999999999999,
+      "loss": 1.9164,
       "step": 3194
     },
     {
       "epoch": 5.112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2516047954559326,
+      "learning_rate": 0.00042464999999999994,
+      "loss": 1.779,
       "step": 3195
     },
     {
       "epoch": 5.1136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.81559419631958,
+      "learning_rate": 0.00042449999999999996,
+      "loss": 1.7815,
       "step": 3196
     },
     {
       "epoch": 5.1152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.766302108764648,
+      "learning_rate": 0.00042435,
+      "loss": 2.1049,
       "step": 3197
     },
     {
       "epoch": 5.1168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.273326873779297,
+      "learning_rate": 0.00042419999999999996,
+      "loss": 2.5158,
       "step": 3198
     },
     {
       "epoch": 5.1184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.278658866882324,
+      "learning_rate": 0.00042405,
+      "loss": 2.0045,
       "step": 3199
     },
     {
       "epoch": 5.12,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 3200
-    },
-    {
-      "epoch": 5.12,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 157.8432,
-      "eval_samples_per_second": 19.868,
-      "eval_steps_per_second": 1.242,
-      "eval_wer": 1.0,
+      "grad_norm": 1.9064677953720093,
+      "learning_rate": 0.00042389999999999995,
+      "loss": 2.1329,
       "step": 3200
     },
     {
       "epoch": 5.1216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0119874477386475,
+      "learning_rate": 0.00042375,
+      "loss": 1.7553,
       "step": 3201
     },
     {
       "epoch": 5.1232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9547804594039917,
+      "learning_rate": 0.00042359999999999994,
+      "loss": 1.8721,
       "step": 3202
     },
     {
       "epoch": 5.1248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.477467656135559,
+      "learning_rate": 0.00042344999999999997,
+      "loss": 1.98,
       "step": 3203
     },
     {
       "epoch": 5.1264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.984140396118164,
+      "learning_rate": 0.0004233,
+      "loss": 2.4288,
       "step": 3204
     },
     {
       "epoch": 5.128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.811163067817688,
+      "learning_rate": 0.00042315,
+      "loss": 2.2058,
       "step": 3205
     },
     {
       "epoch": 5.1296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.115196704864502,
+      "learning_rate": 0.00042299999999999993,
+      "loss": 1.8964,
       "step": 3206
     },
     {
       "epoch": 5.1312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3901374340057373,
+      "learning_rate": 0.00042284999999999995,
+      "loss": 2.0045,
       "step": 3207
     },
     {
       "epoch": 5.1328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.495665550231934,
+      "learning_rate": 0.0004227,
+      "loss": 1.905,
       "step": 3208
     },
     {
       "epoch": 5.1344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.345617294311523,
+      "learning_rate": 0.00042255,
+      "loss": 2.1364,
       "step": 3209
     },
     {
       "epoch": 5.136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.81538987159729,
+      "learning_rate": 0.0004223999999999999,
+      "loss": 2.1264,
       "step": 3210
     },
     {
       "epoch": 5.1376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2732255458831787,
+      "learning_rate": 0.00042224999999999994,
+      "loss": 2.0875,
       "step": 3211
     },
     {
       "epoch": 5.1392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7457600831985474,
+      "learning_rate": 0.00042209999999999996,
+      "loss": 1.9938,
       "step": 3212
     },
     {
       "epoch": 5.1408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7632317543029785,
+      "learning_rate": 0.00042195,
+      "loss": 2.0523,
       "step": 3213
     },
     {
       "epoch": 5.1424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.035547733306885,
+      "learning_rate": 0.00042179999999999995,
+      "loss": 2.2168,
       "step": 3214
     },
     {
       "epoch": 5.144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.828839659690857,
+      "learning_rate": 0.00042165,
+      "loss": 3.0716,
       "step": 3215
     },
     {
       "epoch": 5.1456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.513362407684326,
+      "learning_rate": 0.00042149999999999995,
+      "loss": 2.0964,
       "step": 3216
     },
     {
       "epoch": 5.1472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.751229286193848,
+      "learning_rate": 0.00042134999999999997,
+      "loss": 1.8053,
       "step": 3217
     },
     {
       "epoch": 5.1488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.622213125228882,
+      "learning_rate": 0.00042119999999999994,
+      "loss": 1.9891,
       "step": 3218
     },
     {
       "epoch": 5.1504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1266136169433594,
+      "learning_rate": 0.00042104999999999996,
+      "loss": 2.2839,
       "step": 3219
     },
     {
       "epoch": 5.152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8702733516693115,
+      "learning_rate": 0.0004209,
+      "loss": 1.7316,
       "step": 3220
     },
     {
       "epoch": 5.1536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3464245796203613,
+      "learning_rate": 0.00042075,
+      "loss": 2.4212,
       "step": 3221
     },
     {
       "epoch": 5.1552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9818917512893677,
+      "learning_rate": 0.0004205999999999999,
+      "loss": 2.3367,
       "step": 3222
     },
     {
       "epoch": 5.1568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5446746349334717,
+      "learning_rate": 0.00042044999999999995,
+      "loss": 1.878,
       "step": 3223
     },
     {
       "epoch": 5.1584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.786088943481445,
+      "learning_rate": 0.00042029999999999997,
+      "loss": 2.8948,
       "step": 3224
     },
     {
       "epoch": 5.16,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00042029999999999997,
+      "loss": 2.5519,
       "step": 3225
     },
     {
       "epoch": 5.1616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.343994140625,
+      "learning_rate": 0.00042015,
+      "loss": 2.784,
       "step": 3226
     },
     {
       "epoch": 5.1632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.6946587562561035,
+      "learning_rate": 0.00041999999999999996,
+      "loss": 2.3843,
       "step": 3227
     },
     {
       "epoch": 5.1648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.742859125137329,
+      "learning_rate": 0.00041984999999999993,
+      "loss": 2.1749,
       "step": 3228
     },
     {
       "epoch": 5.1664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.633478164672852,
+      "learning_rate": 0.00041969999999999996,
+      "loss": 2.3875,
       "step": 3229
     },
     {
       "epoch": 5.168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2521109580993652,
+      "learning_rate": 0.00041955,
+      "loss": 2.3516,
       "step": 3230
     },
     {
       "epoch": 5.1696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 21.70244598388672,
+      "learning_rate": 0.00041939999999999995,
+      "loss": 2.8914,
       "step": 3231
     },
     {
       "epoch": 5.1712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.249197244644165,
+      "learning_rate": 0.00041924999999999997,
+      "loss": 2.1809,
       "step": 3232
     },
     {
       "epoch": 5.1728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.294630765914917,
+      "learning_rate": 0.0004191,
+      "loss": 2.3573,
       "step": 3233
     },
     {
       "epoch": 5.1744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.855384349822998,
+      "learning_rate": 0.00041894999999999996,
+      "loss": 2.3276,
       "step": 3234
     },
     {
       "epoch": 5.176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1429007053375244,
+      "learning_rate": 0.00041879999999999993,
+      "loss": 1.8539,
       "step": 3235
     },
     {
       "epoch": 5.1776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8854267597198486,
+      "learning_rate": 0.00041864999999999996,
+      "loss": 2.1204,
       "step": 3236
     },
     {
       "epoch": 5.1792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.373618125915527,
+      "learning_rate": 0.0004185,
+      "loss": 2.8097,
       "step": 3237
     },
     {
       "epoch": 5.1808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.8381149768829346,
+      "learning_rate": 0.00041835,
+      "loss": 1.9282,
       "step": 3238
     },
     {
       "epoch": 5.1824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.196768045425415,
+      "learning_rate": 0.0004181999999999999,
+      "loss": 2.1187,
       "step": 3239
     },
     {
       "epoch": 5.184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.200750350952148,
+      "learning_rate": 0.00041804999999999994,
+      "loss": 1.8194,
       "step": 3240
     },
     {
       "epoch": 5.1856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9149186611175537,
+      "learning_rate": 0.00041789999999999997,
+      "loss": 2.2479,
       "step": 3241
     },
     {
       "epoch": 5.1872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7128161191940308,
+      "learning_rate": 0.00041775,
+      "loss": 1.8903,
       "step": 3242
     },
     {
       "epoch": 5.1888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5912396907806396,
+      "learning_rate": 0.00041759999999999996,
+      "loss": 2.1393,
       "step": 3243
     },
     {
       "epoch": 5.1904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.645106315612793,
+      "learning_rate": 0.00041744999999999993,
+      "loss": 1.9836,
       "step": 3244
     },
     {
       "epoch": 5.192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.632805347442627,
+      "learning_rate": 0.00041729999999999995,
+      "loss": 1.9239,
       "step": 3245
     },
     {
       "epoch": 5.1936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.526568412780762,
+      "learning_rate": 0.00041715,
+      "loss": 2.2065,
       "step": 3246
     },
     {
       "epoch": 5.1952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.276612281799316,
+      "learning_rate": 0.00041699999999999994,
+      "loss": 1.8527,
       "step": 3247
     },
     {
       "epoch": 5.1968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1258394718170166,
+      "learning_rate": 0.00041684999999999997,
+      "loss": 1.4672,
       "step": 3248
     },
     {
       "epoch": 5.1984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9970532655715942,
+      "learning_rate": 0.0004167,
+      "loss": 2.665,
       "step": 3249
     },
     {
       "epoch": 5.2,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5611498355865479,
+      "learning_rate": 0.00041654999999999996,
+      "loss": 2.3542,
       "step": 3250
     },
     {
       "epoch": 5.2016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6552271842956543,
+      "learning_rate": 0.00041639999999999993,
+      "loss": 2.312,
       "step": 3251
     },
     {
       "epoch": 5.2032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3266196250915527,
+      "learning_rate": 0.00041624999999999995,
+      "loss": 1.9171,
       "step": 3252
     },
     {
       "epoch": 5.2048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.908831834793091,
+      "learning_rate": 0.0004161,
+      "loss": 1.6583,
       "step": 3253
     },
     {
       "epoch": 5.2064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.550462245941162,
+      "learning_rate": 0.00041595,
+      "loss": 1.9416,
       "step": 3254
     },
     {
       "epoch": 5.208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.041257381439209,
+      "learning_rate": 0.0004157999999999999,
+      "loss": 1.8542,
       "step": 3255
     },
     {
       "epoch": 5.2096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1703436374664307,
+      "learning_rate": 0.00041564999999999994,
+      "loss": 2.1299,
       "step": 3256
     },
     {
       "epoch": 5.2112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.600370168685913,
+      "learning_rate": 0.00041549999999999996,
+      "loss": 2.3074,
       "step": 3257
     },
     {
       "epoch": 5.2128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4538733959198,
+      "learning_rate": 0.00041535,
+      "loss": 1.8745,
       "step": 3258
     },
     {
       "epoch": 5.2144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.525084018707275,
+      "learning_rate": 0.00041519999999999995,
+      "loss": 2.2919,
       "step": 3259
     },
     {
       "epoch": 5.216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9094107151031494,
+      "learning_rate": 0.0004150499999999999,
+      "loss": 2.9068,
       "step": 3260
     },
     {
       "epoch": 5.2176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6780409812927246,
+      "learning_rate": 0.00041489999999999995,
+      "loss": 2.0547,
       "step": 3261
     },
     {
       "epoch": 5.2192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.290412664413452,
+      "learning_rate": 0.00041474999999999997,
+      "loss": 2.5694,
       "step": 3262
     },
     {
       "epoch": 5.2208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.3367204666137695,
+      "learning_rate": 0.00041459999999999994,
+      "loss": 1.9833,
       "step": 3263
     },
     {
       "epoch": 5.2224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.052272319793701,
+      "learning_rate": 0.00041444999999999996,
+      "loss": 1.9832,
       "step": 3264
     },
     {
       "epoch": 5.224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0450096130371094,
+      "learning_rate": 0.0004143,
+      "loss": 1.9363,
       "step": 3265
     },
     {
       "epoch": 5.2256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9471709728240967,
+      "learning_rate": 0.00041414999999999996,
+      "loss": 2.1991,
       "step": 3266
     },
     {
       "epoch": 5.2272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.396254062652588,
+      "learning_rate": 0.0004139999999999999,
+      "loss": 2.2656,
       "step": 3267
     },
     {
       "epoch": 5.2288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.650744915008545,
+      "learning_rate": 0.00041384999999999995,
+      "loss": 2.6204,
       "step": 3268
     },
     {
       "epoch": 5.2304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.017385959625244,
+      "learning_rate": 0.00041369999999999997,
+      "loss": 2.0147,
       "step": 3269
     },
     {
       "epoch": 5.232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.572188854217529,
+      "learning_rate": 0.00041355,
+      "loss": 2.088,
       "step": 3270
     },
     {
       "epoch": 5.2336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2741122245788574,
+      "learning_rate": 0.0004133999999999999,
+      "loss": 2.1394,
       "step": 3271
     },
     {
       "epoch": 5.2352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6230545043945312,
+      "learning_rate": 0.00041324999999999993,
+      "loss": 2.0355,
       "step": 3272
     },
     {
       "epoch": 5.2368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.208142042160034,
+      "learning_rate": 0.00041309999999999996,
+      "loss": 2.2254,
       "step": 3273
     },
     {
       "epoch": 5.2384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.0993072986602783,
+      "learning_rate": 0.00041295,
+      "loss": 3.0118,
       "step": 3274
     },
     {
       "epoch": 5.24,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9154711961746216,
+      "learning_rate": 0.00041279999999999995,
+      "loss": 3.0291,
       "step": 3275
     },
     {
       "epoch": 5.2416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 24.453245162963867,
+      "learning_rate": 0.00041265,
+      "loss": 3.0981,
       "step": 3276
     },
     {
       "epoch": 5.2432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.712783813476562,
+      "learning_rate": 0.00041249999999999994,
+      "loss": 2.6749,
       "step": 3277
     },
     {
       "epoch": 5.2448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 14.625894546508789,
+      "learning_rate": 0.00041234999999999997,
+      "loss": 3.1377,
       "step": 3278
     },
     {
       "epoch": 5.2464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 19.11041259765625,
+      "learning_rate": 0.0004122,
+      "loss": 3.5501,
       "step": 3279
     },
     {
       "epoch": 5.248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8894535303115845,
+      "learning_rate": 0.00041204999999999996,
+      "loss": 2.2125,
       "step": 3280
     },
     {
       "epoch": 5.2496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0875790119171143,
+      "learning_rate": 0.0004119,
+      "loss": 2.1383,
       "step": 3281
     },
     {
       "epoch": 5.2512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.760862112045288,
+      "learning_rate": 0.00041175,
+      "loss": 2.1078,
       "step": 3282
     },
     {
       "epoch": 5.2528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.318727731704712,
+      "learning_rate": 0.0004116,
+      "loss": 3.0934,
       "step": 3283
     },
     {
       "epoch": 5.2544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.975752830505371,
+      "learning_rate": 0.00041144999999999995,
+      "loss": 2.2509,
       "step": 3284
     },
     {
       "epoch": 5.256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.216862916946411,
+      "learning_rate": 0.00041129999999999997,
+      "loss": 2.1086,
       "step": 3285
     },
     {
       "epoch": 5.2576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.52018404006958,
+      "learning_rate": 0.00041115,
+      "loss": 2.0231,
       "step": 3286
     },
     {
       "epoch": 5.2592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.959451675415039,
+      "learning_rate": 0.000411,
+      "loss": 1.9253,
       "step": 3287
     },
     {
       "epoch": 5.2608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.943277359008789,
+      "learning_rate": 0.00041084999999999993,
+      "loss": 2.074,
       "step": 3288
     },
     {
       "epoch": 5.2624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.6196024417877197,
+      "learning_rate": 0.00041069999999999995,
+      "loss": 1.9088,
       "step": 3289
     },
     {
       "epoch": 5.264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.83423376083374,
+      "learning_rate": 0.00041055,
+      "loss": 2.3001,
       "step": 3290
     },
     {
       "epoch": 5.2656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.978485107421875,
+      "learning_rate": 0.0004104,
+      "loss": 1.9418,
       "step": 3291
     },
     {
       "epoch": 5.2672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1441781520843506,
+      "learning_rate": 0.00041024999999999997,
+      "loss": 2.1912,
       "step": 3292
     },
     {
       "epoch": 5.2688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7500698566436768,
+      "learning_rate": 0.00041009999999999994,
+      "loss": 1.9621,
       "step": 3293
     },
     {
       "epoch": 5.2704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.964430570602417,
+      "learning_rate": 0.00040994999999999996,
+      "loss": 1.9461,
       "step": 3294
     },
     {
       "epoch": 5.272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8824044466018677,
+      "learning_rate": 0.0004098,
+      "loss": 1.7739,
       "step": 3295
     },
     {
       "epoch": 5.2736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.2687458992004395,
+      "learning_rate": 0.00040964999999999996,
+      "loss": 2.1379,
       "step": 3296
     },
     {
       "epoch": 5.2752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.864386558532715,
+      "learning_rate": 0.0004095,
+      "loss": 2.4708,
       "step": 3297
     },
     {
       "epoch": 5.2768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9345788955688477,
+      "learning_rate": 0.00040935,
+      "loss": 1.6851,
       "step": 3298
     },
     {
       "epoch": 5.2783999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8477360010147095,
+      "learning_rate": 0.00040919999999999997,
+      "loss": 2.1697,
       "step": 3299
     },
     {
       "epoch": 5.28,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 3300
-    },
-    {
-      "epoch": 5.28,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 157.0827,
-      "eval_samples_per_second": 19.964,
-      "eval_steps_per_second": 1.248,
-      "eval_wer": 1.0,
+      "grad_norm": 2.215449810028076,
+      "learning_rate": 0.00040904999999999994,
+      "loss": 2.0753,
       "step": 3300
     },
     {
       "epoch": 5.2816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.495771408081055,
+      "learning_rate": 0.00040889999999999996,
+      "loss": 2.2299,
       "step": 3301
     },
     {
       "epoch": 5.2832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.886584520339966,
+      "learning_rate": 0.00040875,
+      "loss": 2.0877,
       "step": 3302
     },
     {
       "epoch": 5.2848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.448219299316406,
+      "learning_rate": 0.0004086,
+      "loss": 2.1397,
       "step": 3303
     },
     {
       "epoch": 5.2864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2363290786743164,
+      "learning_rate": 0.0004084499999999999,
+      "loss": 1.6692,
       "step": 3304
     },
     {
       "epoch": 5.288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9133745431900024,
+      "learning_rate": 0.00040829999999999995,
+      "loss": 2.0965,
       "step": 3305
     },
     {
       "epoch": 5.2896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.235393524169922,
+      "learning_rate": 0.00040815,
+      "loss": 1.719,
       "step": 3306
     },
     {
       "epoch": 5.2912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2992868423461914,
+      "learning_rate": 0.000408,
+      "loss": 1.8493,
       "step": 3307
     },
     {
       "epoch": 5.2928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.721340179443359,
+      "learning_rate": 0.00040784999999999997,
+      "loss": 2.3836,
       "step": 3308
     },
     {
       "epoch": 5.2943999999999996,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3495800495147705,
+      "learning_rate": 0.00040769999999999994,
+      "loss": 2.3834,
       "step": 3309
     },
     {
       "epoch": 5.296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6632156372070312,
+      "learning_rate": 0.00040754999999999996,
+      "loss": 2.0363,
       "step": 3310
     },
     {
       "epoch": 5.2976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5388312339782715,
+      "learning_rate": 0.0004074,
+      "loss": 1.9148,
       "step": 3311
     },
     {
       "epoch": 5.2992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6432439088821411,
+      "learning_rate": 0.00040724999999999995,
+      "loss": 2.4588,
       "step": 3312
     },
     {
       "epoch": 5.3008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.365741729736328,
+      "learning_rate": 0.0004071,
+      "loss": 2.131,
       "step": 3313
     },
     {
       "epoch": 5.3024000000000004,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5189990997314453,
+      "learning_rate": 0.00040695,
+      "loss": 2.0987,
       "step": 3314
     },
     {
       "epoch": 5.304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2603230476379395,
+      "learning_rate": 0.00040679999999999997,
+      "loss": 2.0255,
       "step": 3315
     },
     {
       "epoch": 5.3056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1874077320098877,
+      "learning_rate": 0.00040664999999999994,
+      "loss": 2.3192,
       "step": 3316
     },
     {
       "epoch": 5.3072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8295190334320068,
+      "learning_rate": 0.00040649999999999996,
+      "loss": 2.5205,
       "step": 3317
     },
     {
       "epoch": 5.3088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.067250967025757,
+      "learning_rate": 0.00040635,
+      "loss": 2.5474,
       "step": 3318
     },
     {
       "epoch": 5.3104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.5971553325653076,
+      "learning_rate": 0.0004062,
+      "loss": 2.0922,
       "step": 3319
     },
     {
       "epoch": 5.312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9171907901763916,
+      "learning_rate": 0.0004060499999999999,
+      "loss": 1.9282,
       "step": 3320
     },
     {
       "epoch": 5.3136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9470739364624023,
+      "learning_rate": 0.00040589999999999995,
+      "loss": 1.7039,
       "step": 3321
     },
     {
       "epoch": 5.3152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.5318472385406494,
+      "learning_rate": 0.00040574999999999997,
+      "loss": 2.2137,
       "step": 3322
     },
     {
       "epoch": 5.3168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.8926937580108643,
+      "learning_rate": 0.0004056,
+      "loss": 1.9639,
       "step": 3323
     },
     {
       "epoch": 5.3184000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.356811285018921,
+      "learning_rate": 0.00040544999999999996,
+      "loss": 2.893,
       "step": 3324
     },
     {
       "epoch": 5.32,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1108455657958984,
+      "learning_rate": 0.00040529999999999993,
+      "loss": 2.7859,
       "step": 3325
     },
     {
       "epoch": 5.3216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 11.323676109313965,
+      "learning_rate": 0.00040514999999999995,
+      "loss": 3.1925,
       "step": 3326
     },
     {
       "epoch": 5.3232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.932748794555664,
+      "learning_rate": 0.000405,
+      "loss": 3.6857,
       "step": 3327
     },
     {
       "epoch": 5.3248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.096813917160034,
+      "learning_rate": 0.00040484999999999995,
+      "loss": 2.5022,
       "step": 3328
     },
     {
       "epoch": 5.3264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.869088888168335,
+      "learning_rate": 0.00040469999999999997,
+      "loss": 3.3066,
       "step": 3329
     },
     {
       "epoch": 5.328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.629360198974609,
+      "learning_rate": 0.00040455,
+      "loss": 2.8505,
       "step": 3330
     },
     {
       "epoch": 5.3296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.395150661468506,
+      "learning_rate": 0.0004044,
+      "loss": 2.4143,
       "step": 3331
     },
     {
       "epoch": 5.3312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4230806827545166,
+      "learning_rate": 0.00040424999999999993,
+      "loss": 2.1592,
       "step": 3332
     },
     {
       "epoch": 5.3328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8740898370742798,
+      "learning_rate": 0.00040409999999999996,
+      "loss": 2.2058,
       "step": 3333
     },
     {
       "epoch": 5.3344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7961093187332153,
+      "learning_rate": 0.00040395,
+      "loss": 2.0633,
       "step": 3334
     },
     {
       "epoch": 5.336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.026645183563232,
+      "learning_rate": 0.0004038,
+      "loss": 2.2895,
       "step": 3335
     },
     {
       "epoch": 5.3376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 16.91562843322754,
+      "learning_rate": 0.0004036499999999999,
+      "loss": 2.7079,
       "step": 3336
     },
     {
       "epoch": 5.3392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.92512321472168,
+      "learning_rate": 0.00040349999999999994,
+      "loss": 2.4684,
       "step": 3337
     },
     {
       "epoch": 5.3408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.841024398803711,
+      "learning_rate": 0.00040334999999999997,
+      "loss": 2.1537,
       "step": 3338
     },
     {
       "epoch": 5.3424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.517487525939941,
+      "learning_rate": 0.0004032,
+      "loss": 2.3195,
       "step": 3339
     },
     {
       "epoch": 5.344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.065592765808105,
+      "learning_rate": 0.00040304999999999996,
+      "loss": 2.2745,
       "step": 3340
     },
     {
       "epoch": 5.3456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.872082233428955,
+      "learning_rate": 0.0004029,
+      "loss": 2.0909,
       "step": 3341
     },
     {
       "epoch": 5.3472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.373950242996216,
+      "learning_rate": 0.00040274999999999995,
+      "loss": 2.1205,
       "step": 3342
     },
     {
       "epoch": 5.3488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.094278335571289,
+      "learning_rate": 0.0004026,
+      "loss": 1.5954,
       "step": 3343
     },
     {
       "epoch": 5.3504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.5784664154052734,
+      "learning_rate": 0.00040244999999999994,
+      "loss": 2.2211,
       "step": 3344
     },
     {
       "epoch": 5.352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7646541595458984,
+      "learning_rate": 0.00040229999999999997,
+      "loss": 2.3838,
       "step": 3345
     },
     {
       "epoch": 5.3536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.288341760635376,
+      "learning_rate": 0.00040215,
+      "loss": 2.14,
       "step": 3346
     },
     {
       "epoch": 5.3552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.88338565826416,
+      "learning_rate": 0.000402,
+      "loss": 2.1205,
       "step": 3347
     },
     {
       "epoch": 5.3568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1948492527008057,
+      "learning_rate": 0.00040184999999999993,
+      "loss": 1.8892,
       "step": 3348
     },
     {
       "epoch": 5.3584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.734545707702637,
+      "learning_rate": 0.00040169999999999995,
+      "loss": 2.4243,
       "step": 3349
     },
     {
       "epoch": 5.36,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.707056045532227,
+      "learning_rate": 0.00040155,
+      "loss": 2.1569,
       "step": 3350
     },
     {
       "epoch": 5.3616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.175364971160889,
+      "learning_rate": 0.0004014,
+      "loss": 2.235,
       "step": 3351
     },
     {
       "epoch": 5.3632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.4466357231140137,
+      "learning_rate": 0.0004012499999999999,
+      "loss": 2.2742,
       "step": 3352
     },
     {
       "epoch": 5.3648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7935688495635986,
+      "learning_rate": 0.00040109999999999994,
+      "loss": 1.9754,
       "step": 3353
     },
     {
       "epoch": 5.3664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.9915544986724854,
+      "learning_rate": 0.00040094999999999996,
+      "loss": 1.8166,
       "step": 3354
     },
     {
       "epoch": 5.368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.6275420188903809,
+      "learning_rate": 0.0004008,
+      "loss": 2.0787,
       "step": 3355
     },
     {
       "epoch": 5.3696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.187203884124756,
+      "learning_rate": 0.00040064999999999995,
+      "loss": 2.2121,
       "step": 3356
     },
     {
       "epoch": 5.3712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.500221014022827,
+      "learning_rate": 0.0004005,
+      "loss": 2.1276,
       "step": 3357
     },
     {
       "epoch": 5.3728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.457571983337402,
+      "learning_rate": 0.00040034999999999995,
+      "loss": 2.1685,
       "step": 3358
     },
     {
       "epoch": 5.3744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.376873016357422,
+      "learning_rate": 0.00040019999999999997,
+      "loss": 1.9855,
       "step": 3359
     },
     {
       "epoch": 5.376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6002018451690674,
+      "learning_rate": 0.00040004999999999994,
+      "loss": 2.1904,
       "step": 3360
     },
     {
       "epoch": 5.3776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.4764018058776855,
+      "learning_rate": 0.00039989999999999996,
+      "loss": 2.3107,
       "step": 3361
     },
     {
       "epoch": 5.3792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1630468368530273,
+      "learning_rate": 0.00039975,
+      "loss": 2.4739,
       "step": 3362
     },
     {
       "epoch": 5.3808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.243349552154541,
+      "learning_rate": 0.0003996,
+      "loss": 2.2559,
       "step": 3363
     },
     {
       "epoch": 5.3824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1403260231018066,
+      "learning_rate": 0.0003994499999999999,
+      "loss": 2.0141,
       "step": 3364
     },
     {
       "epoch": 5.384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.301318883895874,
+      "learning_rate": 0.00039929999999999995,
+      "loss": 2.1598,
       "step": 3365
     },
     {
       "epoch": 5.3856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.172241687774658,
+      "learning_rate": 0.00039914999999999997,
+      "loss": 2.2622,
       "step": 3366
     },
     {
       "epoch": 5.3872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.561786413192749,
+      "learning_rate": 0.000399,
+      "loss": 2.2528,
       "step": 3367
     },
     {
       "epoch": 5.3888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.7395148277282715,
+      "learning_rate": 0.0003988499999999999,
+      "loss": 2.0952,
       "step": 3368
     },
     {
       "epoch": 5.3904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.103436470031738,
+      "learning_rate": 0.00039869999999999993,
+      "loss": 2.119,
       "step": 3369
     },
     {
       "epoch": 5.392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.751752853393555,
+      "learning_rate": 0.00039854999999999996,
+      "loss": 2.4361,
       "step": 3370
     },
     {
       "epoch": 5.3936,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00039854999999999996,
+      "loss": 1.1379,
       "step": 3371
     },
     {
       "epoch": 5.3952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6950693130493164,
+      "learning_rate": 0.0003984,
+      "loss": 2.5055,
       "step": 3372
     },
     {
       "epoch": 5.3968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.234592914581299,
+      "learning_rate": 0.00039824999999999995,
+      "loss": 2.8761,
       "step": 3373
     },
     {
       "epoch": 5.3984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.145230770111084,
+      "learning_rate": 0.0003981,
+      "loss": 2.7371,
       "step": 3374
     },
     {
       "epoch": 5.4,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.621272087097168,
+      "learning_rate": 0.00039794999999999994,
+      "loss": 1.9957,
       "step": 3375
     },
     {
       "epoch": 5.4016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 24.97957420349121,
+      "learning_rate": 0.00039779999999999997,
+      "loss": 3.3891,
       "step": 3376
     },
     {
       "epoch": 5.4032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 16.701343536376953,
+      "learning_rate": 0.00039764999999999993,
+      "loss": 2.8649,
       "step": 3377
     },
     {
       "epoch": 5.4048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.291032791137695,
+      "learning_rate": 0.00039749999999999996,
+      "loss": 2.7527,
       "step": 3378
     },
     {
       "epoch": 5.4064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 19.17793083190918,
+      "learning_rate": 0.00039735,
+      "loss": 3.1116,
       "step": 3379
     },
     {
       "epoch": 5.408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.143411636352539,
+      "learning_rate": 0.0003972,
+      "loss": 2.7099,
       "step": 3380
     },
     {
       "epoch": 5.4096,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.2232489585876465,
+      "learning_rate": 0.0003970499999999999,
+      "loss": 2.1449,
       "step": 3381
     },
     {
       "epoch": 5.4112,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.380960464477539,
+      "learning_rate": 0.00039689999999999994,
+      "loss": 2.5124,
       "step": 3382
     },
     {
       "epoch": 5.4128,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.133636474609375,
+      "learning_rate": 0.00039674999999999997,
+      "loss": 2.6039,
       "step": 3383
     },
     {
       "epoch": 5.4144,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.810513973236084,
+      "learning_rate": 0.0003966,
+      "loss": 2.3203,
       "step": 3384
     },
     {
       "epoch": 5.416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.090081214904785,
+      "learning_rate": 0.00039644999999999996,
+      "loss": 2.132,
       "step": 3385
     },
     {
       "epoch": 5.4176,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.350454330444336,
+      "learning_rate": 0.00039629999999999993,
+      "loss": 2.3344,
       "step": 3386
     },
     {
       "epoch": 5.4192,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 11.359246253967285,
+      "learning_rate": 0.00039614999999999995,
+      "loss": 2.8971,
       "step": 3387
     },
     {
       "epoch": 5.4208,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.318723678588867,
+      "learning_rate": 0.000396,
+      "loss": 2.6054,
       "step": 3388
     },
     {
       "epoch": 5.4224,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 13.670693397521973,
+      "learning_rate": 0.00039584999999999995,
+      "loss": 2.8112,
       "step": 3389
     },
     {
       "epoch": 5.424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.342107772827148,
+      "learning_rate": 0.00039569999999999997,
+      "loss": 2.1677,
       "step": 3390
     },
     {
       "epoch": 5.4256,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.924628257751465,
+      "learning_rate": 0.00039555,
+      "loss": 2.0552,
       "step": 3391
     },
     {
       "epoch": 5.4272,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.202759742736816,
+      "learning_rate": 0.00039539999999999996,
+      "loss": 2.5917,
       "step": 3392
     },
     {
       "epoch": 5.4288,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.981536388397217,
+      "learning_rate": 0.00039524999999999993,
+      "loss": 2.2385,
       "step": 3393
     },
     {
       "epoch": 5.4304,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.929697513580322,
+      "learning_rate": 0.00039509999999999995,
+      "loss": 2.6301,
       "step": 3394
     },
     {
       "epoch": 5.432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.891195774078369,
+      "learning_rate": 0.00039495,
+      "loss": 2.068,
       "step": 3395
     },
     {
       "epoch": 5.4336,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2171473503112793,
+      "learning_rate": 0.0003948,
+      "loss": 2.0446,
       "step": 3396
     },
     {
       "epoch": 5.4352,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.248323440551758,
+      "learning_rate": 0.0003946499999999999,
+      "loss": 2.7933,
       "step": 3397
     },
     {
       "epoch": 5.4368,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1067569255828857,
+      "learning_rate": 0.00039449999999999994,
+      "loss": 2.3266,
       "step": 3398
     },
     {
       "epoch": 5.4384,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.4660804271698,
+      "learning_rate": 0.00039434999999999996,
+      "loss": 1.958,
       "step": 3399
     },
     {
       "epoch": 5.44,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
-      "step": 3400
-    },
-    {
-      "epoch": 5.44,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 156.8368,
-      "eval_samples_per_second": 19.995,
-      "eval_steps_per_second": 1.25,
-      "eval_wer": 1.0,
+      "grad_norm": 2.094003438949585,
+      "learning_rate": 0.0003942,
+      "loss": 1.9817,
       "step": 3400
     },
     {
       "epoch": 5.4416,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.3582420349121094,
+      "learning_rate": 0.00039404999999999996,
+      "loss": 2.4704,
       "step": 3401
     },
     {
       "epoch": 5.4432,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.880425930023193,
+      "learning_rate": 0.0003938999999999999,
+      "loss": 2.3661,
       "step": 3402
     },
     {
       "epoch": 5.4448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.494797706604004,
+      "learning_rate": 0.00039374999999999995,
+      "loss": 2.3579,
       "step": 3403
     },
     {
       "epoch": 5.4464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.800097465515137,
+      "learning_rate": 0.00039359999999999997,
+      "loss": 2.2806,
       "step": 3404
     },
     {
       "epoch": 5.448,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.084938049316406,
+      "learning_rate": 0.00039345,
+      "loss": 2.2602,
       "step": 3405
     },
     {
       "epoch": 5.4496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.3108110427856445,
+      "learning_rate": 0.00039329999999999996,
+      "loss": 2.6064,
       "step": 3406
     },
     {
       "epoch": 5.4512,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.094543218612671,
+      "learning_rate": 0.00039315,
+      "loss": 2.3647,
       "step": 3407
     },
     {
       "epoch": 5.4528,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6065499782562256,
+      "learning_rate": 0.00039299999999999996,
+      "loss": 2.1223,
       "step": 3408
     },
     {
       "epoch": 5.4544,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.361059665679932,
+      "learning_rate": 0.00039285,
+      "loss": 2.1135,
       "step": 3409
     },
     {
       "epoch": 5.456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1946380138397217,
+      "learning_rate": 0.00039269999999999995,
+      "loss": 2.2807,
       "step": 3410
     },
     {
       "epoch": 5.4576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.014134407043457,
+      "learning_rate": 0.00039255,
+      "loss": 2.1623,
       "step": 3411
     },
     {
       "epoch": 5.4592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.53760290145874,
+      "learning_rate": 0.0003924,
+      "loss": 2.2694,
       "step": 3412
     },
     {
       "epoch": 5.4608,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.014392614364624,
+      "learning_rate": 0.00039225,
+      "loss": 2.0437,
       "step": 3413
     },
     {
       "epoch": 5.4624,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.662132501602173,
+      "learning_rate": 0.00039209999999999994,
+      "loss": 2.2526,
       "step": 3414
     },
     {
       "epoch": 5.464,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.099125623703003,
+      "learning_rate": 0.00039194999999999996,
+      "loss": 2.0935,
       "step": 3415
     },
     {
       "epoch": 5.4656,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.434375762939453,
+      "learning_rate": 0.0003918,
+      "loss": 2.3328,
       "step": 3416
     },
     {
       "epoch": 5.4672,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.892804145812988,
+      "learning_rate": 0.00039165,
+      "loss": 2.2687,
       "step": 3417
     },
     {
       "epoch": 5.4688,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.694879531860352,
+      "learning_rate": 0.0003914999999999999,
+      "loss": 2.2931,
       "step": 3418
     },
     {
       "epoch": 5.4704,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.8057279586792,
+      "learning_rate": 0.00039134999999999994,
+      "loss": 2.3536,
       "step": 3419
     },
     {
       "epoch": 5.4719999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 15.479071617126465,
+      "learning_rate": 0.00039119999999999997,
+      "loss": 3.7015,
       "step": 3420
     },
     {
       "epoch": 5.4736,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.293941974639893,
+      "learning_rate": 0.00039105,
+      "loss": 2.3939,
       "step": 3421
     },
     {
       "epoch": 5.4752,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.1204912662506104,
+      "learning_rate": 0.00039089999999999996,
+      "loss": 2.637,
       "step": 3422
     },
     {
       "epoch": 5.4768,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.834355354309082,
+      "learning_rate": 0.00039075,
+      "loss": 1.9505,
       "step": 3423
     },
     {
       "epoch": 5.4784,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9132513999938965,
+      "learning_rate": 0.00039059999999999995,
+      "loss": 2.3591,
       "step": 3424
     },
     {
       "epoch": 5.48,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00039059999999999995,
+      "loss": 1.9456,
       "step": 3425
     },
     {
       "epoch": 5.4816,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 30.53275489807129,
+      "learning_rate": 0.00039045,
+      "loss": 3.8161,
       "step": 3426
     },
     {
       "epoch": 5.4832,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 10.212854385375977,
+      "learning_rate": 0.00039029999999999995,
+      "loss": 2.9696,
       "step": 3427
     },
     {
       "epoch": 5.4848,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.351596832275391,
+      "learning_rate": 0.00039014999999999997,
+      "loss": 2.5098,
       "step": 3428
     },
     {
       "epoch": 5.4864,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6440775394439697,
+      "learning_rate": 0.00039,
+      "loss": 2.3737,
       "step": 3429
     },
     {
       "epoch": 5.4879999999999995,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7896642088890076,
+      "learning_rate": 0.00038985,
+      "loss": 2.6933,
       "step": 3430
     },
     {
       "epoch": 5.4896,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.894845008850098,
+      "learning_rate": 0.00038969999999999993,
+      "loss": 3.1879,
       "step": 3431
     },
     {
       "epoch": 5.4912,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.497279644012451,
+      "learning_rate": 0.00038954999999999995,
+      "loss": 3.2892,
       "step": 3432
     },
     {
       "epoch": 5.4928,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.8545002937316895,
+      "learning_rate": 0.0003894,
+      "loss": 2.6857,
       "step": 3433
     },
     {
       "epoch": 5.4944,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.6429190635681152,
+      "learning_rate": 0.00038925,
+      "loss": 2.8884,
       "step": 3434
     },
     {
       "epoch": 5.496,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.311533451080322,
+      "learning_rate": 0.00038909999999999997,
+      "loss": 2.7945,
       "step": 3435
     },
     {
       "epoch": 5.4976,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.832476615905762,
+      "learning_rate": 0.00038894999999999994,
+      "loss": 3.3598,
       "step": 3436
     },
     {
       "epoch": 5.4992,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.438668727874756,
+      "learning_rate": 0.00038879999999999996,
+      "loss": 2.9759,
       "step": 3437
     },
     {
       "epoch": 5.5008,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.911144733428955,
+      "learning_rate": 0.00038865,
+      "loss": 2.8816,
       "step": 3438
     },
     {
       "epoch": 5.5024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.37494945526123,
+      "learning_rate": 0.00038849999999999996,
+      "loss": 2.6733,
       "step": 3439
     },
     {
       "epoch": 5.504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.502228736877441,
+      "learning_rate": 0.00038835,
+      "loss": 2.7077,
       "step": 3440
     },
     {
       "epoch": 5.5056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.751497268676758,
+      "learning_rate": 0.0003882,
+      "loss": 2.3219,
       "step": 3441
     },
     {
       "epoch": 5.5072,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8348023891448975,
+      "learning_rate": 0.00038804999999999997,
+      "loss": 2.2012,
       "step": 3442
     },
     {
       "epoch": 5.5088,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.414168119430542,
+      "learning_rate": 0.00038789999999999994,
+      "loss": 2.2623,
       "step": 3443
     },
     {
       "epoch": 5.5104,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.748080253601074,
+      "learning_rate": 0.00038774999999999997,
+      "loss": 2.4969,
       "step": 3444
     },
     {
       "epoch": 5.5120000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.4296817779541016,
+      "learning_rate": 0.0003876,
+      "loss": 1.9788,
       "step": 3445
     },
     {
       "epoch": 5.5136,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 9.921895027160645,
+      "learning_rate": 0.00038745,
+      "loss": 2.5149,
       "step": 3446
     },
     {
       "epoch": 5.5152,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.484654426574707,
+      "learning_rate": 0.00038729999999999993,
+      "loss": 1.9065,
       "step": 3447
     },
     {
       "epoch": 5.5168,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.423427104949951,
+      "learning_rate": 0.00038714999999999995,
+      "loss": 2.3022,
       "step": 3448
     },
     {
       "epoch": 5.5184,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.991278171539307,
+      "learning_rate": 0.000387,
+      "loss": 2.2629,
       "step": 3449
     },
     {
       "epoch": 5.52,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.440671443939209,
+      "learning_rate": 0.00038685,
+      "loss": 2.3491,
       "step": 3450
     },
     {
       "epoch": 5.5216,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.21368408203125,
+      "learning_rate": 0.00038669999999999997,
+      "loss": 2.3761,
       "step": 3451
     },
     {
       "epoch": 5.5232,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.914760112762451,
+      "learning_rate": 0.00038654999999999994,
+      "loss": 1.9627,
       "step": 3452
     },
     {
       "epoch": 5.5248,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 5.628223419189453,
+      "learning_rate": 0.00038639999999999996,
+      "loss": 2.0523,
       "step": 3453
     },
     {
       "epoch": 5.5264,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 6.52740478515625,
+      "learning_rate": 0.00038625,
+      "loss": 2.0522,
       "step": 3454
     },
     {
       "epoch": 5.5280000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8526406288146973,
+      "learning_rate": 0.00038609999999999995,
+      "loss": 2.3311,
       "step": 3455
     },
     {
       "epoch": 5.5296,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.2300243377685547,
+      "learning_rate": 0.00038595,
+      "loss": 2.1947,
       "step": 3456
     },
     {
       "epoch": 5.5312,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.7482554912567139,
+      "learning_rate": 0.0003858,
+      "loss": 2.5054,
       "step": 3457
     },
     {
       "epoch": 5.5328,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.598001003265381,
+      "learning_rate": 0.00038564999999999997,
+      "loss": 3.2034,
       "step": 3458
     },
     {
       "epoch": 5.5344,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.464871644973755,
+      "learning_rate": 0.00038549999999999994,
+      "loss": 1.5426,
       "step": 3459
     },
     {
       "epoch": 5.536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.8509607315063477,
+      "learning_rate": 0.00038534999999999996,
+      "loss": 1.5835,
       "step": 3460
     },
     {
       "epoch": 5.5376,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.15891170501709,
+      "learning_rate": 0.0003852,
+      "loss": 2.5372,
       "step": 3461
     },
     {
       "epoch": 5.5392,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.5455851554870605,
+      "learning_rate": 0.00038505,
+      "loss": 2.0308,
       "step": 3462
     },
     {
       "epoch": 5.5408,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.959388494491577,
+      "learning_rate": 0.0003848999999999999,
+      "loss": 2.235,
       "step": 3463
     },
     {
       "epoch": 5.5424,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.463346481323242,
+      "learning_rate": 0.00038474999999999995,
+      "loss": 2.6703,
       "step": 3464
     },
     {
       "epoch": 5.5440000000000005,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 7.2167463302612305,
+      "learning_rate": 0.00038459999999999997,
+      "loss": 2.4555,
       "step": 3465
     },
     {
       "epoch": 5.5456,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.825636148452759,
+      "learning_rate": 0.00038445,
+      "loss": 2.6392,
       "step": 3466
     },
     {
       "epoch": 5.5472,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 14.59807300567627,
+      "learning_rate": 0.00038429999999999996,
+      "loss": 2.4381,
       "step": 3467
     },
     {
       "epoch": 5.5488,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.1133155822753906,
+      "learning_rate": 0.00038414999999999993,
+      "loss": 2.6275,
       "step": 3468
     },
     {
       "epoch": 5.5504,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 4.662886142730713,
+      "learning_rate": 0.00038399999999999996,
+      "loss": 2.5494,
       "step": 3469
     },
     {
       "epoch": 5.552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.9676240682601929,
+      "learning_rate": 0.00038385,
+      "loss": 1.9896,
       "step": 3470
     },
     {
       "epoch": 5.5536,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.780965805053711,
+      "learning_rate": 0.00038369999999999995,
+      "loss": 2.4608,
       "step": 3471
     },
     {
       "epoch": 5.5552,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 3.9176764488220215,
+      "learning_rate": 0.00038354999999999997,
+      "loss": 3.1066,
       "step": 3472
     },
     {
       "epoch": 5.5568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.076561450958252,
+      "learning_rate": 0.0003834,
+      "loss": 2.5004,
       "step": 3473
     },
     {
       "epoch": 5.5584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.902611255645752,
+      "learning_rate": 0.00038324999999999996,
+      "loss": 2.4972,
       "step": 3474
     },
     {
       "epoch": 5.5600000000000005,
       "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "learning_rate": 0.00038324999999999996,
+      "loss": 2.5864,
       "step": 3475
     },
     {
       "epoch": 5.5616,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 23.75803565979004,
+      "learning_rate": 0.00038309999999999993,
+      "loss": 3.8795,
       "step": 3476
     },
     {
       "epoch": 5.5632,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.7207627296447754,
+      "learning_rate": 0.00038294999999999996,
+      "loss": 4.0748,
       "step": 3477
     },
     {
       "epoch": 5.5648,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 2.0137767791748047,
+      "learning_rate": 0.0003828,
+      "loss": 2.3535,
       "step": 3478
     },
     {
       "epoch": 5.5664,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.7027060389518738,
+      "learning_rate": 0.00038265,
+      "loss": 3.2328,
       "step": 3479
     },
     {
       "epoch": 5.568,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.26203009486198425,
+      "learning_rate": 0.0003824999999999999,
+      "loss": 2.3086,
       "step": 3480
     },
     {
       "epoch": 5.5696,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.36936789751052856,
+      "learning_rate": 0.00038234999999999994,
+      "loss": 3.1161,
       "step": 3481
     },
     {
       "epoch": 5.5712,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 8.606059074401855,
+      "learning_rate": 0.00038219999999999997,
+      "loss": 2.9592,
       "step": 3482
     },
     {
       "epoch": 5.5728,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.1602068841457367,
+      "learning_rate": 0.00038205,
+      "loss": 3.157,
       "step": 3483
     },
     {
       "epoch": 5.5744,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.08841034024953842,
+      "learning_rate": 0.00038189999999999996,
+      "loss": 2.4829,
       "step": 3484
     },
     {
       "epoch": 5.576,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5950487852096558,
+      "learning_rate": 0.00038175,
+      "loss": 2.1481,
       "step": 3485
     },
     {
       "epoch": 5.5776,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.12550248205661774,
+      "learning_rate": 0.00038159999999999995,
+      "loss": 2.4758,
       "step": 3486
     },
     {
       "epoch": 5.5792,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.06971815228462219,
+      "learning_rate": 0.00038145,
+      "loss": 2.541,
       "step": 3487
     },
     {
       "epoch": 5.5808,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4338502585887909,
+      "learning_rate": 0.00038129999999999994,
+      "loss": 2.9192,
       "step": 3488
     },
     {
       "epoch": 5.5824,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00038114999999999997,
+      "loss": 2.3753,
       "step": 3489
     },
     {
       "epoch": 5.584,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.1862034946680069,
+      "learning_rate": 0.000381,
+      "loss": 2.557,
       "step": 3490
     },
     {
       "epoch": 5.5856,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.25080394744873047,
+      "learning_rate": 0.00038085,
+      "loss": 3.3535,
       "step": 3491
     },
     {
       "epoch": 5.5872,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.13513095676898956,
+      "learning_rate": 0.00038069999999999993,
+      "loss": 3.3913,
       "step": 3492
     },
     {
       "epoch": 5.5888,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.5417481064796448,
+      "learning_rate": 0.00038054999999999995,
+      "loss": 2.5585,
       "step": 3493
     },
     {
       "epoch": 5.5904,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.21010702848434448,
+      "learning_rate": 0.0003804,
+      "loss": 2.7133,
       "step": 3494
     },
     {
       "epoch": 5.592,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.4073018431663513,
+      "learning_rate": 0.00038025,
+      "loss": 2.8361,
       "step": 3495
     },
     {
       "epoch": 5.5936,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.24190618097782135,
+      "learning_rate": 0.0003800999999999999,
+      "loss": 2.6598,
       "step": 3496
     },
     {
       "epoch": 5.5952,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.671675980091095,
+      "learning_rate": 0.00037994999999999994,
+      "loss": 2.7558,
       "step": 3497
     },
     {
       "epoch": 5.5968,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.17936047911643982,
+      "learning_rate": 0.00037979999999999996,
+      "loss": 2.7637,
       "step": 3498
     },
     {
       "epoch": 5.5984,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 0.3960895538330078,
+      "learning_rate": 0.00037965,
+      "loss": 2.969,
       "step": 3499
     },
     {
       "epoch": 5.6,
-      "grad_norm": NaN,
-      "learning_rate": 0.0005752,
-      "loss": 0.0,
+      "grad_norm": 1.2539879083633423,
+      "learning_rate": 0.00037949999999999995,
+      "loss": 3.0468,
       "step": 3500
     },
     {
-      "epoch": 5.6,
-      "eval_cer": 1.0,
-      "eval_loss": NaN,
-      "eval_runtime": 157.361,
-      "eval_samples_per_second": 19.929,
-      "eval_steps_per_second": 1.246,
-      "eval_wer": 1.0,
-      "step": 3500
+      "epoch": 5.6016,
+      "grad_norm": 0.6254233717918396,
+      "learning_rate": 0.00037935,
+      "loss": 2.3933,
+      "step": 3501
     },
     {
-      "epoch": 5.6,
-      "step": 3500,
-      "total_flos": 2.158464150901847e+19,
-      "train_loss": 1.3790143143790108,
-      "train_runtime": 12492.843,
-      "train_samples_per_second": 8.965,
-      "train_steps_per_second": 0.28
+      "epoch": 5.6032,
+      "grad_norm": 0.4243174195289612,
+      "learning_rate": 0.00037919999999999995,
+      "loss": 2.109,
+      "step": 3502
+    },
+    {
+      "epoch": 5.6048,
+      "grad_norm": 0.5422927737236023,
+      "learning_rate": 0.00037904999999999997,
+      "loss": 1.9256,
+      "step": 3503
+    },
+    {
+      "epoch": 5.6064,
+      "grad_norm": 0.2607356011867523,
+      "learning_rate": 0.00037889999999999994,
+      "loss": 2.259,
+      "step": 3504
+    },
+    {
+      "epoch": 5.608,
+      "grad_norm": 2.16225528717041,
+      "learning_rate": 0.00037874999999999996,
+      "loss": 2.3937,
+      "step": 3505
+    },
+    {
+      "epoch": 5.6096,
+      "grad_norm": 0.2814367115497589,
+      "learning_rate": 0.0003786,
+      "loss": 1.9617,
+      "step": 3506
+    },
+    {
+      "epoch": 5.6112,
+      "grad_norm": 1.0939712524414062,
+      "learning_rate": 0.00037845,
+      "loss": 2.1647,
+      "step": 3507
+    },
+    {
+      "epoch": 5.6128,
+      "grad_norm": 1.2848657369613647,
+      "learning_rate": 0.0003782999999999999,
+      "loss": 2.2722,
+      "step": 3508
+    },
+    {
+      "epoch": 5.6144,
+      "grad_norm": 2.4309964179992676,
+      "learning_rate": 0.00037814999999999995,
+      "loss": 2.2048,
+      "step": 3509
+    },
+    {
+      "epoch": 5.616,
+      "grad_norm": 1.734381079673767,
+      "learning_rate": 0.00037799999999999997,
+      "loss": 2.2176,
+      "step": 3510
+    },
+    {
+      "epoch": 5.6176,
+      "grad_norm": 0.7754119038581848,
+      "learning_rate": 0.00037785,
+      "loss": 1.9389,
+      "step": 3511
+    },
+    {
+      "epoch": 5.6192,
+      "grad_norm": 1.9875555038452148,
+      "learning_rate": 0.0003776999999999999,
+      "loss": 2.1693,
+      "step": 3512
+    },
+    {
+      "epoch": 5.6208,
+      "grad_norm": 1.4995112419128418,
+      "learning_rate": 0.00037754999999999993,
+      "loss": 2.4732,
+      "step": 3513
+    },
+    {
+      "epoch": 5.6224,
+      "grad_norm": 1.3808419704437256,
+      "learning_rate": 0.00037739999999999996,
+      "loss": 2.8265,
+      "step": 3514
+    },
+    {
+      "epoch": 5.624,
+      "grad_norm": 0.639369010925293,
+      "learning_rate": 0.00037725,
+      "loss": 2.1947,
+      "step": 3515
+    },
+    {
+      "epoch": 5.6256,
+      "grad_norm": 4.914621353149414,
+      "learning_rate": 0.00037709999999999995,
+      "loss": 3.165,
+      "step": 3516
+    },
+    {
+      "epoch": 5.6272,
+      "grad_norm": 1.8339734077453613,
+      "learning_rate": 0.00037695,
+      "loss": 2.7099,
+      "step": 3517
+    },
+    {
+      "epoch": 5.6288,
+      "grad_norm": 1.6919299364089966,
+      "learning_rate": 0.00037679999999999994,
+      "loss": 2.301,
+      "step": 3518
+    },
+    {
+      "epoch": 5.6304,
+      "grad_norm": 2.5645647048950195,
+      "learning_rate": 0.00037664999999999997,
+      "loss": 2.4862,
+      "step": 3519
+    },
+    {
+      "epoch": 5.632,
+      "grad_norm": 3.0452866554260254,
+      "learning_rate": 0.00037649999999999994,
+      "loss": 2.4736,
+      "step": 3520
+    },
+    {
+      "epoch": 5.6336,
+      "grad_norm": 1.4942389726638794,
+      "learning_rate": 0.00037634999999999996,
+      "loss": 1.8172,
+      "step": 3521
+    },
+    {
+      "epoch": 5.6352,
+      "grad_norm": 1.7136120796203613,
+      "learning_rate": 0.0003762,
+      "loss": 2.9583,
+      "step": 3522
+    },
+    {
+      "epoch": 5.6368,
+      "grad_norm": 2.3411545753479004,
+      "learning_rate": 0.00037605,
+      "loss": 3.5991,
+      "step": 3523
+    },
+    {
+      "epoch": 5.6384,
+      "grad_norm": 5.684845924377441,
+      "learning_rate": 0.0003758999999999999,
+      "loss": 3.5991,
+      "step": 3524
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 8.414681434631348,
+      "learning_rate": 0.00037574999999999994,
+      "loss": 4.8554,
+      "step": 3525
+    },
+    {
+      "epoch": 5.6416,
+      "grad_norm": 23.659269332885742,
+      "learning_rate": 0.00037559999999999997,
+      "loss": 3.7371,
+      "step": 3526
+    },
+    {
+      "epoch": 5.6432,
+      "grad_norm": 10.667974472045898,
+      "learning_rate": 0.00037545,
+      "loss": 3.066,
+      "step": 3527
+    },
+    {
+      "epoch": 5.6448,
+      "grad_norm": 2.8220784664154053,
+      "learning_rate": 0.00037529999999999996,
+      "loss": 3.5351,
+      "step": 3528
+    },
+    {
+      "epoch": 5.6464,
+      "grad_norm": 3.089872121810913,
+      "learning_rate": 0.00037514999999999993,
+      "loss": 2.5887,
+      "step": 3529
+    },
+    {
+      "epoch": 5.648,
+      "grad_norm": 1.261480689048767,
+      "learning_rate": 0.00037499999999999995,
+      "loss": 2.7372,
+      "step": 3530
+    },
+    {
+      "epoch": 5.6495999999999995,
+      "grad_norm": 0.26717910170555115,
+      "learning_rate": 0.00037485,
+      "loss": 2.5624,
+      "step": 3531
+    },
+    {
+      "epoch": 5.6512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003747,
+      "loss": 3.0436,
+      "step": 3532
+    },
+    {
+      "epoch": 5.6528,
+      "grad_norm": 0.7398553490638733,
+      "learning_rate": 0.00037454999999999997,
+      "loss": 2.7866,
+      "step": 3533
+    },
+    {
+      "epoch": 5.6544,
+      "grad_norm": 0.6313406229019165,
+      "learning_rate": 0.0003744,
+      "loss": 2.9158,
+      "step": 3534
+    },
+    {
+      "epoch": 5.656,
+      "grad_norm": 1.2333641052246094,
+      "learning_rate": 0.00037424999999999996,
+      "loss": 3.5131,
+      "step": 3535
+    },
+    {
+      "epoch": 5.6576,
+      "grad_norm": 1.7500239610671997,
+      "learning_rate": 0.0003741,
+      "loss": 3.7137,
+      "step": 3536
+    },
+    {
+      "epoch": 5.6592,
+      "grad_norm": 0.331387460231781,
+      "learning_rate": 0.00037394999999999996,
+      "loss": 3.2079,
+      "step": 3537
+    },
+    {
+      "epoch": 5.6608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003738,
+      "loss": 2.5893,
+      "step": 3538
+    },
+    {
+      "epoch": 5.6624,
+      "grad_norm": 0.2625195384025574,
+      "learning_rate": 0.00037365,
+      "loss": 3.4963,
+      "step": 3539
+    },
+    {
+      "epoch": 5.664,
+      "grad_norm": 0.7120262384414673,
+      "learning_rate": 0.0003735,
+      "loss": 3.0325,
+      "step": 3540
+    },
+    {
+      "epoch": 5.6655999999999995,
+      "grad_norm": 0.2676083445549011,
+      "learning_rate": 0.00037334999999999994,
+      "loss": 3.2996,
+      "step": 3541
+    },
+    {
+      "epoch": 5.6672,
+      "grad_norm": 0.14166611433029175,
+      "learning_rate": 0.00037319999999999996,
+      "loss": 3.5891,
+      "step": 3542
+    },
+    {
+      "epoch": 5.6688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00037305,
+      "loss": 2.954,
+      "step": 3543
+    },
+    {
+      "epoch": 5.6704,
+      "grad_norm": 2.4091098308563232,
+      "learning_rate": 0.0003729,
+      "loss": 3.8448,
+      "step": 3544
+    },
+    {
+      "epoch": 5.672,
+      "grad_norm": 4.258470058441162,
+      "learning_rate": 0.0003727499999999999,
+      "loss": 3.1852,
+      "step": 3545
+    },
+    {
+      "epoch": 5.6736,
+      "grad_norm": 0.7352797985076904,
+      "learning_rate": 0.00037259999999999995,
+      "loss": 3.0091,
+      "step": 3546
+    },
+    {
+      "epoch": 5.6752,
+      "grad_norm": 1.294426679611206,
+      "learning_rate": 0.00037245,
+      "loss": 3.116,
+      "step": 3547
+    },
+    {
+      "epoch": 5.6768,
+      "grad_norm": 1.2180641889572144,
+      "learning_rate": 0.0003723,
+      "loss": 3.3635,
+      "step": 3548
+    },
+    {
+      "epoch": 5.6784,
+      "grad_norm": 1.7434768676757812,
+      "learning_rate": 0.00037214999999999997,
+      "loss": 3.1209,
+      "step": 3549
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.20002809166908264,
+      "learning_rate": 0.000372,
+      "loss": 2.7624,
+      "step": 3550
+    },
+    {
+      "epoch": 5.6815999999999995,
+      "grad_norm": 0.7545135617256165,
+      "learning_rate": 0.00037184999999999996,
+      "loss": 2.5356,
+      "step": 3551
+    },
+    {
+      "epoch": 5.6832,
+      "grad_norm": 0.5763547420501709,
+      "learning_rate": 0.0003717,
+      "loss": 2.8453,
+      "step": 3552
+    },
+    {
+      "epoch": 5.6848,
+      "grad_norm": 0.379525363445282,
+      "learning_rate": 0.00037154999999999995,
+      "loss": 2.6852,
+      "step": 3553
+    },
+    {
+      "epoch": 5.6864,
+      "grad_norm": 0.5061140656471252,
+      "learning_rate": 0.0003714,
+      "loss": 2.3777,
+      "step": 3554
+    },
+    {
+      "epoch": 5.688,
+      "grad_norm": 0.7044528126716614,
+      "learning_rate": 0.00037125,
+      "loss": 2.4686,
+      "step": 3555
+    },
+    {
+      "epoch": 5.6896,
+      "grad_norm": 0.6468807458877563,
+      "learning_rate": 0.0003711,
+      "loss": 2.8022,
+      "step": 3556
+    },
+    {
+      "epoch": 5.6912,
+      "grad_norm": 1.682919979095459,
+      "learning_rate": 0.00037094999999999994,
+      "loss": 2.5016,
+      "step": 3557
+    },
+    {
+      "epoch": 5.6928,
+      "grad_norm": 3.406224012374878,
+      "learning_rate": 0.00037079999999999996,
+      "loss": 2.8903,
+      "step": 3558
+    },
+    {
+      "epoch": 5.6944,
+      "grad_norm": 0.8235665559768677,
+      "learning_rate": 0.00037065,
+      "loss": 2.4151,
+      "step": 3559
+    },
+    {
+      "epoch": 5.696,
+      "grad_norm": 4.509483814239502,
+      "learning_rate": 0.0003705,
+      "loss": 2.5733,
+      "step": 3560
+    },
+    {
+      "epoch": 5.6975999999999996,
+      "grad_norm": 0.9885743856430054,
+      "learning_rate": 0.0003703499999999999,
+      "loss": 2.2135,
+      "step": 3561
+    },
+    {
+      "epoch": 5.6992,
+      "grad_norm": 2.305175304412842,
+      "learning_rate": 0.00037019999999999995,
+      "loss": 2.3394,
+      "step": 3562
+    },
+    {
+      "epoch": 5.7008,
+      "grad_norm": 2.077007293701172,
+      "learning_rate": 0.00037004999999999997,
+      "loss": 2.4202,
+      "step": 3563
+    },
+    {
+      "epoch": 5.7024,
+      "grad_norm": 0.6968153715133667,
+      "learning_rate": 0.0003699,
+      "loss": 2.2318,
+      "step": 3564
+    },
+    {
+      "epoch": 5.704,
+      "grad_norm": 2.7181806564331055,
+      "learning_rate": 0.00036974999999999996,
+      "loss": 2.3018,
+      "step": 3565
+    },
+    {
+      "epoch": 5.7056000000000004,
+      "grad_norm": 2.1074941158294678,
+      "learning_rate": 0.0003696,
+      "loss": 3.078,
+      "step": 3566
+    },
+    {
+      "epoch": 5.7072,
+      "grad_norm": 1.7561454772949219,
+      "learning_rate": 0.00036944999999999995,
+      "loss": 1.7743,
+      "step": 3567
+    },
+    {
+      "epoch": 5.7088,
+      "grad_norm": 1.8946216106414795,
+      "learning_rate": 0.0003693,
+      "loss": 2.3119,
+      "step": 3568
+    },
+    {
+      "epoch": 5.7104,
+      "grad_norm": 1.7721720933914185,
+      "learning_rate": 0.00036914999999999995,
+      "loss": 2.3399,
+      "step": 3569
+    },
+    {
+      "epoch": 5.712,
+      "grad_norm": 2.348559617996216,
+      "learning_rate": 0.00036899999999999997,
+      "loss": 2.8106,
+      "step": 3570
+    },
+    {
+      "epoch": 5.7136,
+      "grad_norm": 3.1461150646209717,
+      "learning_rate": 0.00036885,
+      "loss": 2.2077,
+      "step": 3571
+    },
+    {
+      "epoch": 5.7152,
+      "grad_norm": 2.274102210998535,
+      "learning_rate": 0.0003687,
+      "loss": 2.4829,
+      "step": 3572
+    },
+    {
+      "epoch": 5.7168,
+      "grad_norm": 2.2827303409576416,
+      "learning_rate": 0.00036854999999999993,
+      "loss": 3.999,
+      "step": 3573
+    },
+    {
+      "epoch": 5.7184,
+      "grad_norm": 2.7197787761688232,
+      "learning_rate": 0.00036839999999999996,
+      "loss": 2.9606,
+      "step": 3574
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 2.6948750019073486,
+      "learning_rate": 0.00036825,
+      "loss": 3.1148,
+      "step": 3575
+    },
+    {
+      "epoch": 5.7216000000000005,
+      "grad_norm": 0.272156298160553,
+      "learning_rate": 0.0003681,
+      "loss": 2.705,
+      "step": 3576
+    },
+    {
+      "epoch": 5.7232,
+      "grad_norm": 3.451700448989868,
+      "learning_rate": 0.0003679499999999999,
+      "loss": 3.3249,
+      "step": 3577
+    },
+    {
+      "epoch": 5.7248,
+      "grad_norm": 3.6281368732452393,
+      "learning_rate": 0.00036779999999999994,
+      "loss": 3.0924,
+      "step": 3578
+    },
+    {
+      "epoch": 5.7264,
+      "grad_norm": 10.167502403259277,
+      "learning_rate": 0.00036764999999999996,
+      "loss": 3.0343,
+      "step": 3579
+    },
+    {
+      "epoch": 5.728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003675,
+      "loss": 2.8255,
+      "step": 3580
+    },
+    {
+      "epoch": 5.7296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00036734999999999996,
+      "loss": 3.0536,
+      "step": 3581
+    },
+    {
+      "epoch": 5.7312,
+      "grad_norm": 7.263051509857178,
+      "learning_rate": 0.0003672,
+      "loss": 3.2911,
+      "step": 3582
+    },
+    {
+      "epoch": 5.7328,
+      "grad_norm": 0.07337801903486252,
+      "learning_rate": 0.00036705,
+      "loss": 2.583,
+      "step": 3583
+    },
+    {
+      "epoch": 5.7344,
+      "grad_norm": 3.855539321899414,
+      "learning_rate": 0.0003669,
+      "loss": 2.5903,
+      "step": 3584
+    },
+    {
+      "epoch": 5.736,
+      "grad_norm": 0.20969152450561523,
+      "learning_rate": 0.00036674999999999994,
+      "loss": 2.6072,
+      "step": 3585
+    },
+    {
+      "epoch": 5.7376000000000005,
+      "grad_norm": 1.757079839706421,
+      "learning_rate": 0.00036659999999999997,
+      "loss": 2.3773,
+      "step": 3586
+    },
+    {
+      "epoch": 5.7392,
+      "grad_norm": 4.414907932281494,
+      "learning_rate": 0.00036645,
+      "loss": 2.956,
+      "step": 3587
+    },
+    {
+      "epoch": 5.7408,
+      "grad_norm": 0.1729482263326645,
+      "learning_rate": 0.0003663,
+      "loss": 2.659,
+      "step": 3588
+    },
+    {
+      "epoch": 5.7424,
+      "grad_norm": 0.10146791487932205,
+      "learning_rate": 0.00036614999999999993,
+      "loss": 2.9967,
+      "step": 3589
+    },
+    {
+      "epoch": 5.744,
+      "grad_norm": 0.11373867094516754,
+      "learning_rate": 0.00036599999999999995,
+      "loss": 2.4121,
+      "step": 3590
+    },
+    {
+      "epoch": 5.7456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00036585,
+      "loss": 2.4116,
+      "step": 3591
+    },
+    {
+      "epoch": 5.7472,
+      "grad_norm": 0.2513590455055237,
+      "learning_rate": 0.0003657,
+      "loss": 2.8289,
+      "step": 3592
+    },
+    {
+      "epoch": 5.7488,
+      "grad_norm": 0.7403852343559265,
+      "learning_rate": 0.00036554999999999997,
+      "loss": 2.9567,
+      "step": 3593
+    },
+    {
+      "epoch": 5.7504,
+      "grad_norm": 0.31389737129211426,
+      "learning_rate": 0.00036539999999999994,
+      "loss": 2.6591,
+      "step": 3594
+    },
+    {
+      "epoch": 5.752,
+      "grad_norm": 0.8933272957801819,
+      "learning_rate": 0.00036524999999999996,
+      "loss": 2.7838,
+      "step": 3595
+    },
+    {
+      "epoch": 5.7536000000000005,
+      "grad_norm": 1.0159043073654175,
+      "learning_rate": 0.0003651,
+      "loss": 2.7835,
+      "step": 3596
+    },
+    {
+      "epoch": 5.7552,
+      "grad_norm": 0.21672067046165466,
+      "learning_rate": 0.00036494999999999995,
+      "loss": 3.1109,
+      "step": 3597
+    },
+    {
+      "epoch": 5.7568,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0003648,
+      "loss": 2.7365,
+      "step": 3598
+    },
+    {
+      "epoch": 5.7584,
+      "grad_norm": 0.7378922700881958,
+      "learning_rate": 0.00036465,
+      "loss": 3.2703,
+      "step": 3599
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 2.0105292797088623,
+      "learning_rate": 0.00036449999999999997,
+      "loss": 3.349,
+      "step": 3600
+    },
+    {
+      "epoch": 5.7616,
+      "grad_norm": 1.5307704210281372,
+      "learning_rate": 0.00036434999999999994,
+      "loss": 2.8438,
+      "step": 3601
+    },
+    {
+      "epoch": 5.7632,
+      "grad_norm": 1.4970020055770874,
+      "learning_rate": 0.00036419999999999996,
+      "loss": 3.2691,
+      "step": 3602
+    },
+    {
+      "epoch": 5.7648,
+      "grad_norm": 0.9958275556564331,
+      "learning_rate": 0.00036405,
+      "loss": 2.8416,
+      "step": 3603
+    },
+    {
+      "epoch": 5.7664,
+      "grad_norm": 0.8194541931152344,
+      "learning_rate": 0.0003639,
+      "loss": 2.6063,
+      "step": 3604
+    },
+    {
+      "epoch": 5.768,
+      "grad_norm": 1.198455572128296,
+      "learning_rate": 0.0003637499999999999,
+      "loss": 3.0273,
+      "step": 3605
+    },
+    {
+      "epoch": 5.7696,
+      "grad_norm": 0.8404093384742737,
+      "learning_rate": 0.00036359999999999995,
+      "loss": 2.5248,
+      "step": 3606
+    },
+    {
+      "epoch": 5.7712,
+      "grad_norm": 0.1331993043422699,
+      "learning_rate": 0.00036344999999999997,
+      "loss": 2.1761,
+      "step": 3607
+    },
+    {
+      "epoch": 5.7728,
+      "grad_norm": 1.9269537925720215,
+      "learning_rate": 0.0003633,
+      "loss": 2.9078,
+      "step": 3608
+    },
+    {
+      "epoch": 5.7744,
+      "grad_norm": 2.229436159133911,
+      "learning_rate": 0.00036314999999999996,
+      "loss": 2.3954,
+      "step": 3609
+    },
+    {
+      "epoch": 5.776,
+      "grad_norm": 1.4860639572143555,
+      "learning_rate": 0.00036299999999999993,
+      "loss": 2.3718,
+      "step": 3610
+    },
+    {
+      "epoch": 5.7776,
+      "grad_norm": 1.9439291954040527,
+      "learning_rate": 0.00036284999999999996,
+      "loss": 3.4277,
+      "step": 3611
+    },
+    {
+      "epoch": 5.7792,
+      "grad_norm": 1.416441798210144,
+      "learning_rate": 0.0003627,
+      "loss": 2.1693,
+      "step": 3612
+    },
+    {
+      "epoch": 5.7808,
+      "grad_norm": 1.086266279220581,
+      "learning_rate": 0.00036254999999999995,
+      "loss": 2.3808,
+      "step": 3613
+    },
+    {
+      "epoch": 5.7824,
+      "grad_norm": 0.8500319719314575,
+      "learning_rate": 0.00036239999999999997,
+      "loss": 2.603,
+      "step": 3614
+    },
+    {
+      "epoch": 5.784,
+      "grad_norm": 2.137688398361206,
+      "learning_rate": 0.00036225,
+      "loss": 2.4094,
+      "step": 3615
+    },
+    {
+      "epoch": 5.7856,
+      "grad_norm": 1.138893961906433,
+      "learning_rate": 0.00036209999999999997,
+      "loss": 2.5953,
+      "step": 3616
+    },
+    {
+      "epoch": 5.7872,
+      "grad_norm": 2.8222107887268066,
+      "learning_rate": 0.00036194999999999993,
+      "loss": 2.2547,
+      "step": 3617
+    },
+    {
+      "epoch": 5.7888,
+      "grad_norm": 1.6344839334487915,
+      "learning_rate": 0.00036179999999999996,
+      "loss": 2.3833,
+      "step": 3618
+    },
+    {
+      "epoch": 5.7904,
+      "grad_norm": 1.5543922185897827,
+      "learning_rate": 0.00036165,
+      "loss": 2.8417,
+      "step": 3619
+    },
+    {
+      "epoch": 5.792,
+      "grad_norm": 1.5485179424285889,
+      "learning_rate": 0.0003615,
+      "loss": 2.4023,
+      "step": 3620
+    },
+    {
+      "epoch": 5.7936,
+      "grad_norm": 1.6071698665618896,
+      "learning_rate": 0.0003613499999999999,
+      "loss": 2.8954,
+      "step": 3621
+    },
+    {
+      "epoch": 5.7952,
+      "grad_norm": 1.7881412506103516,
+      "learning_rate": 0.00036119999999999994,
+      "loss": 2.5974,
+      "step": 3622
+    },
+    {
+      "epoch": 5.7968,
+      "grad_norm": 3.4796767234802246,
+      "learning_rate": 0.00036104999999999997,
+      "loss": 2.6164,
+      "step": 3623
+    },
+    {
+      "epoch": 5.7984,
+      "grad_norm": 3.5168821811676025,
+      "learning_rate": 0.0003609,
+      "loss": 2.58,
+      "step": 3624
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 2.84468150138855,
+      "learning_rate": 0.00036074999999999996,
+      "loss": 3.1729,
+      "step": 3625
+    },
+    {
+      "epoch": 5.8016,
+      "grad_norm": 27.352258682250977,
+      "learning_rate": 0.00036059999999999993,
+      "loss": 5.0092,
+      "step": 3626
+    },
+    {
+      "epoch": 5.8032,
+      "grad_norm": 7.655608654022217,
+      "learning_rate": 0.00036044999999999995,
+      "loss": 3.3475,
+      "step": 3627
+    },
+    {
+      "epoch": 5.8048,
+      "grad_norm": 1.47767972946167,
+      "learning_rate": 0.0003603,
+      "loss": 3.133,
+      "step": 3628
+    },
+    {
+      "epoch": 5.8064,
+      "grad_norm": 0.7489213347434998,
+      "learning_rate": 0.00036014999999999995,
+      "loss": 2.7145,
+      "step": 3629
+    },
+    {
+      "epoch": 5.808,
+      "grad_norm": 9.507126808166504,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 2.7299,
+      "step": 3630
+    },
+    {
+      "epoch": 5.8096,
+      "grad_norm": 1.0871667861938477,
+      "learning_rate": 0.00035985,
+      "loss": 3.3557,
+      "step": 3631
+    },
+    {
+      "epoch": 5.8112,
+      "grad_norm": 0.6891421675682068,
+      "learning_rate": 0.00035969999999999996,
+      "loss": 2.453,
+      "step": 3632
+    },
+    {
+      "epoch": 5.8128,
+      "grad_norm": 0.11214147508144379,
+      "learning_rate": 0.00035954999999999993,
+      "loss": 2.9922,
+      "step": 3633
+    },
+    {
+      "epoch": 5.8144,
+      "grad_norm": 0.45183080434799194,
+      "learning_rate": 0.00035939999999999995,
+      "loss": 3.4575,
+      "step": 3634
+    },
+    {
+      "epoch": 5.816,
+      "grad_norm": 0.42742919921875,
+      "learning_rate": 0.00035925,
+      "loss": 2.6807,
+      "step": 3635
+    },
+    {
+      "epoch": 5.8176,
+      "grad_norm": 0.3587559163570404,
+      "learning_rate": 0.0003591,
+      "loss": 2.689,
+      "step": 3636
+    },
+    {
+      "epoch": 5.8192,
+      "grad_norm": 0.12587009370326996,
+      "learning_rate": 0.0003589499999999999,
+      "loss": 2.9673,
+      "step": 3637
+    },
+    {
+      "epoch": 5.8208,
+      "grad_norm": 0.16796007752418518,
+      "learning_rate": 0.00035879999999999994,
+      "loss": 2.9272,
+      "step": 3638
+    },
+    {
+      "epoch": 5.8224,
+      "grad_norm": 0.8416668176651001,
+      "learning_rate": 0.00035864999999999996,
+      "loss": 3.1985,
+      "step": 3639
+    },
+    {
+      "epoch": 5.824,
+      "grad_norm": 1.0569196939468384,
+      "learning_rate": 0.0003585,
+      "loss": 3.362,
+      "step": 3640
+    },
+    {
+      "epoch": 5.8256,
+      "grad_norm": 0.45992952585220337,
+      "learning_rate": 0.00035834999999999996,
+      "loss": 3.0646,
+      "step": 3641
+    },
+    {
+      "epoch": 5.8272,
+      "grad_norm": 0.04979749768972397,
+      "learning_rate": 0.0003582,
+      "loss": 2.7772,
+      "step": 3642
+    },
+    {
+      "epoch": 5.8288,
+      "grad_norm": 2.914752960205078,
+      "learning_rate": 0.00035804999999999995,
+      "loss": 3.0556,
+      "step": 3643
+    },
+    {
+      "epoch": 5.8304,
+      "grad_norm": 0.5485931634902954,
+      "learning_rate": 0.00035789999999999997,
+      "loss": 2.7492,
+      "step": 3644
+    },
+    {
+      "epoch": 5.832,
+      "grad_norm": 0.425957053899765,
+      "learning_rate": 0.00035774999999999994,
+      "loss": 2.737,
+      "step": 3645
+    },
+    {
+      "epoch": 5.8336,
+      "grad_norm": 0.3038780987262726,
+      "learning_rate": 0.00035759999999999996,
+      "loss": 3.1534,
+      "step": 3646
+    },
+    {
+      "epoch": 5.8352,
+      "grad_norm": 0.6664896011352539,
+      "learning_rate": 0.00035745,
+      "loss": 2.8512,
+      "step": 3647
+    },
+    {
+      "epoch": 5.8368,
+      "grad_norm": 0.5831376910209656,
+      "learning_rate": 0.0003573,
+      "loss": 2.8303,
+      "step": 3648
+    },
+    {
+      "epoch": 5.8384,
+      "grad_norm": 0.7792990803718567,
+      "learning_rate": 0.0003571499999999999,
+      "loss": 2.8147,
+      "step": 3649
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.7054243683815002,
+      "learning_rate": 0.00035699999999999995,
+      "loss": 2.6731,
+      "step": 3650
+    },
+    {
+      "epoch": 5.8416,
+      "grad_norm": 0.44943201541900635,
+      "learning_rate": 0.00035685,
+      "loss": 2.6966,
+      "step": 3651
+    },
+    {
+      "epoch": 5.8431999999999995,
+      "grad_norm": 0.7449806928634644,
+      "learning_rate": 0.0003567,
+      "loss": 2.7262,
+      "step": 3652
+    },
+    {
+      "epoch": 5.8448,
+      "grad_norm": 0.5415761470794678,
+      "learning_rate": 0.0003565499999999999,
+      "loss": 2.1463,
+      "step": 3653
+    },
+    {
+      "epoch": 5.8464,
+      "grad_norm": 0.5867460370063782,
+      "learning_rate": 0.00035639999999999994,
+      "loss": 2.2366,
+      "step": 3654
+    },
+    {
+      "epoch": 5.848,
+      "grad_norm": 0.15858949720859528,
+      "learning_rate": 0.00035624999999999996,
+      "loss": 2.8666,
+      "step": 3655
+    },
+    {
+      "epoch": 5.8496,
+      "grad_norm": 1.2337722778320312,
+      "learning_rate": 0.0003561,
+      "loss": 2.2596,
+      "step": 3656
+    },
+    {
+      "epoch": 5.8512,
+      "grad_norm": 0.874224066734314,
+      "learning_rate": 0.00035595,
+      "loss": 2.4329,
+      "step": 3657
+    },
+    {
+      "epoch": 5.8528,
+      "grad_norm": 1.4948076009750366,
+      "learning_rate": 0.0003558,
+      "loss": 3.2546,
+      "step": 3658
+    },
+    {
+      "epoch": 5.8544,
+      "grad_norm": 2.395164728164673,
+      "learning_rate": 0.00035564999999999994,
+      "loss": 2.2181,
+      "step": 3659
+    },
+    {
+      "epoch": 5.856,
+      "grad_norm": 0.8242649435997009,
+      "learning_rate": 0.00035549999999999997,
+      "loss": 2.1852,
+      "step": 3660
+    },
+    {
+      "epoch": 5.8576,
+      "grad_norm": 2.2238521575927734,
+      "learning_rate": 0.00035535,
+      "loss": 3.2672,
+      "step": 3661
+    },
+    {
+      "epoch": 5.8591999999999995,
+      "grad_norm": 0.8653325438499451,
+      "learning_rate": 0.00035519999999999996,
+      "loss": 2.174,
+      "step": 3662
+    },
+    {
+      "epoch": 5.8608,
+      "grad_norm": 1.0872834920883179,
+      "learning_rate": 0.00035505,
+      "loss": 2.316,
+      "step": 3663
+    },
+    {
+      "epoch": 5.8624,
+      "grad_norm": 1.8807929754257202,
+      "learning_rate": 0.0003549,
+      "loss": 2.8189,
+      "step": 3664
+    },
+    {
+      "epoch": 5.864,
+      "grad_norm": 1.446363925933838,
+      "learning_rate": 0.00035475,
+      "loss": 2.4262,
+      "step": 3665
+    },
+    {
+      "epoch": 5.8656,
+      "grad_norm": 2.251727819442749,
+      "learning_rate": 0.00035459999999999995,
+      "loss": 2.5645,
+      "step": 3666
+    },
+    {
+      "epoch": 5.8672,
+      "grad_norm": 1.9902764558792114,
+      "learning_rate": 0.00035444999999999997,
+      "loss": 2.4165,
+      "step": 3667
+    },
+    {
+      "epoch": 5.8688,
+      "grad_norm": 1.670762062072754,
+      "learning_rate": 0.0003543,
+      "loss": 2.6425,
+      "step": 3668
+    },
+    {
+      "epoch": 5.8704,
+      "grad_norm": 8.156818389892578,
+      "learning_rate": 0.00035415,
+      "loss": 2.1805,
+      "step": 3669
+    },
+    {
+      "epoch": 5.872,
+      "grad_norm": 1.7316776514053345,
+      "learning_rate": 0.00035399999999999993,
+      "loss": 2.6375,
+      "step": 3670
+    },
+    {
+      "epoch": 5.8736,
+      "grad_norm": 3.32999587059021,
+      "learning_rate": 0.00035384999999999995,
+      "loss": 2.8535,
+      "step": 3671
+    },
+    {
+      "epoch": 5.8751999999999995,
+      "grad_norm": 1.9965382814407349,
+      "learning_rate": 0.0003537,
+      "loss": 2.7322,
+      "step": 3672
+    },
+    {
+      "epoch": 5.8768,
+      "grad_norm": 2.408020257949829,
+      "learning_rate": 0.00035355,
+      "loss": 2.3558,
+      "step": 3673
+    },
+    {
+      "epoch": 5.8784,
+      "grad_norm": 1.9624814987182617,
+      "learning_rate": 0.00035339999999999997,
+      "loss": 2.6974,
+      "step": 3674
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": NaN,
+      "learning_rate": 0.00035339999999999997,
+      "loss": 2.638,
+      "step": 3675
+    },
+    {
+      "epoch": 5.8816,
+      "grad_norm": 0.5264666676521301,
+      "learning_rate": 0.00035324999999999994,
+      "loss": 3.2748,
+      "step": 3676
+    },
+    {
+      "epoch": 5.8832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035309999999999996,
+      "loss": 3.4704,
+      "step": 3677
+    },
+    {
+      "epoch": 5.8848,
+      "grad_norm": 6.098517894744873,
+      "learning_rate": 0.00035295,
+      "loss": 3.4468,
+      "step": 3678
+    },
+    {
+      "epoch": 5.8864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035279999999999996,
+      "loss": 3.5018,
+      "step": 3679
+    },
+    {
+      "epoch": 5.888,
+      "grad_norm": 0.74627685546875,
+      "learning_rate": 0.00035265,
+      "loss": 2.7214,
+      "step": 3680
+    },
+    {
+      "epoch": 5.8896,
+      "grad_norm": 9.594038963317871,
+      "learning_rate": 0.0003525,
+      "loss": 4.2435,
+      "step": 3681
+    },
+    {
+      "epoch": 5.8911999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035234999999999997,
+      "loss": 3.0054,
+      "step": 3682
+    },
+    {
+      "epoch": 5.8928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035219999999999994,
+      "loss": 3.3124,
+      "step": 3683
+    },
+    {
+      "epoch": 5.8944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035204999999999996,
+      "loss": 2.5724,
+      "step": 3684
+    },
+    {
+      "epoch": 5.896,
+      "grad_norm": 1.967746615409851,
+      "learning_rate": 0.0003519,
+      "loss": 3.026,
+      "step": 3685
+    },
+    {
+      "epoch": 5.8976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035175,
+      "loss": 2.5008,
+      "step": 3686
+    },
+    {
+      "epoch": 5.8992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003515999999999999,
+      "loss": 2.7721,
+      "step": 3687
+    },
+    {
+      "epoch": 5.9008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035144999999999995,
+      "loss": 2.6447,
+      "step": 3688
+    },
+    {
+      "epoch": 5.9024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003513,
+      "loss": 2.1775,
+      "step": 3689
+    },
+    {
+      "epoch": 5.904,
+      "grad_norm": 5.827794551849365,
+      "learning_rate": 0.00035115,
+      "loss": 2.6886,
+      "step": 3690
+    },
+    {
+      "epoch": 5.9056,
+      "grad_norm": 0.3083203434944153,
+      "learning_rate": 0.00035099999999999997,
+      "loss": 2.7191,
+      "step": 3691
+    },
+    {
+      "epoch": 5.9072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035085,
+      "loss": 2.4568,
+      "step": 3692
+    },
+    {
+      "epoch": 5.9088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035069999999999996,
+      "loss": 2.2762,
+      "step": 3693
+    },
+    {
+      "epoch": 5.9104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035055,
+      "loss": 2.6573,
+      "step": 3694
+    },
+    {
+      "epoch": 5.912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035039999999999995,
+      "loss": 2.92,
+      "step": 3695
+    },
+    {
+      "epoch": 5.9136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00035025,
+      "loss": 2.3955,
+      "step": 3696
+    },
+    {
+      "epoch": 5.9152000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003501,
+      "loss": 2.8471,
+      "step": 3697
+    },
+    {
+      "epoch": 5.9168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034995,
+      "loss": 2.3392,
+      "step": 3698
+    },
+    {
+      "epoch": 5.9184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034979999999999994,
+      "loss": 2.4649,
+      "step": 3699
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034964999999999996,
+      "loss": 2.9227,
+      "step": 3700
+    },
+    {
+      "epoch": 5.9216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003495,
+      "loss": 2.2749,
+      "step": 3701
+    },
+    {
+      "epoch": 5.9232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034935,
+      "loss": 2.2452,
+      "step": 3702
+    },
+    {
+      "epoch": 5.9248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003491999999999999,
+      "loss": 2.3583,
+      "step": 3703
+    },
+    {
+      "epoch": 5.9264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034904999999999995,
+      "loss": 2.7241,
+      "step": 3704
+    },
+    {
+      "epoch": 5.928,
+      "grad_norm": 0.5584777593612671,
+      "learning_rate": 0.00034889999999999997,
+      "loss": 2.6595,
+      "step": 3705
+    },
+    {
+      "epoch": 5.9296,
+      "grad_norm": 0.4647485315799713,
+      "learning_rate": 0.00034875,
+      "loss": 3.1368,
+      "step": 3706
+    },
+    {
+      "epoch": 5.9312000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034859999999999996,
+      "loss": 2.6697,
+      "step": 3707
+    },
+    {
+      "epoch": 5.9328,
+      "grad_norm": 0.32228565216064453,
+      "learning_rate": 0.00034845,
+      "loss": 3.4953,
+      "step": 3708
+    },
+    {
+      "epoch": 5.9344,
+      "grad_norm": 0.7425710558891296,
+      "learning_rate": 0.00034829999999999996,
+      "loss": 2.7956,
+      "step": 3709
+    },
+    {
+      "epoch": 5.936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034815,
+      "loss": 2.9307,
+      "step": 3710
+    },
+    {
+      "epoch": 5.9376,
+      "grad_norm": 1.3797106742858887,
+      "learning_rate": 0.00034799999999999995,
+      "loss": 3.6561,
+      "step": 3711
+    },
+    {
+      "epoch": 5.9392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034784999999999997,
+      "loss": 2.8536,
+      "step": 3712
+    },
+    {
+      "epoch": 5.9408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003477,
+      "loss": 3.0708,
+      "step": 3713
+    },
+    {
+      "epoch": 5.9424,
+      "grad_norm": 0.6898555755615234,
+      "learning_rate": 0.00034755,
+      "loss": 2.8878,
+      "step": 3714
+    },
+    {
+      "epoch": 5.944,
+      "grad_norm": 0.3775753080844879,
+      "learning_rate": 0.00034739999999999993,
+      "loss": 2.4076,
+      "step": 3715
+    },
+    {
+      "epoch": 5.9456,
+      "grad_norm": 1.3761353492736816,
+      "learning_rate": 0.00034724999999999996,
+      "loss": 4.4482,
+      "step": 3716
+    },
+    {
+      "epoch": 5.9472000000000005,
+      "grad_norm": 1.383660912513733,
+      "learning_rate": 0.0003471,
+      "loss": 3.4384,
+      "step": 3717
+    },
+    {
+      "epoch": 5.9488,
+      "grad_norm": 0.6061517000198364,
+      "learning_rate": 0.00034695,
+      "loss": 3.4682,
+      "step": 3718
+    },
+    {
+      "epoch": 5.9504,
+      "grad_norm": 0.4879741072654724,
+      "learning_rate": 0.0003467999999999999,
+      "loss": 2.7721,
+      "step": 3719
+    },
+    {
+      "epoch": 5.952,
+      "grad_norm": 0.30795058608055115,
+      "learning_rate": 0.00034664999999999994,
+      "loss": 2.9665,
+      "step": 3720
+    },
+    {
+      "epoch": 5.9536,
+      "grad_norm": 0.7795411944389343,
+      "learning_rate": 0.00034649999999999997,
+      "loss": 3.0061,
+      "step": 3721
+    },
+    {
+      "epoch": 5.9552,
+      "grad_norm": 0.41862058639526367,
+      "learning_rate": 0.00034635,
+      "loss": 3.0164,
+      "step": 3722
+    },
+    {
+      "epoch": 5.9568,
+      "grad_norm": 0.9294652342796326,
+      "learning_rate": 0.00034619999999999996,
+      "loss": 3.2634,
+      "step": 3723
+    },
+    {
+      "epoch": 5.9584,
+      "grad_norm": 1.671874761581421,
+      "learning_rate": 0.00034605,
+      "loss": 2.8787,
+      "step": 3724
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.3849936723709106,
+      "learning_rate": 0.00034589999999999995,
+      "loss": 3.0871,
+      "step": 3725
+    },
+    {
+      "epoch": 5.9616,
+      "grad_norm": 2.8756697177886963,
+      "learning_rate": 0.00034575,
+      "loss": 3.8418,
+      "step": 3726
+    },
+    {
+      "epoch": 5.9632,
+      "grad_norm": 5.069089412689209,
+      "learning_rate": 0.00034559999999999994,
+      "loss": 3.3647,
+      "step": 3727
+    },
+    {
+      "epoch": 5.9648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034544999999999997,
+      "loss": 3.2391,
+      "step": 3728
+    },
+    {
+      "epoch": 5.9664,
+      "grad_norm": 0.6909729242324829,
+      "learning_rate": 0.0003453,
+      "loss": 3.0748,
+      "step": 3729
+    },
+    {
+      "epoch": 5.968,
+      "grad_norm": 0.24212801456451416,
+      "learning_rate": 0.00034515,
+      "loss": 3.2098,
+      "step": 3730
+    },
+    {
+      "epoch": 5.9696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034499999999999993,
+      "loss": 3.4712,
+      "step": 3731
+    },
+    {
+      "epoch": 5.9712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034484999999999995,
+      "loss": 2.919,
+      "step": 3732
+    },
+    {
+      "epoch": 5.9728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003447,
+      "loss": 2.4222,
+      "step": 3733
+    },
+    {
+      "epoch": 5.9744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034455,
+      "loss": 2.9553,
+      "step": 3734
+    },
+    {
+      "epoch": 5.976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034439999999999997,
+      "loss": 2.2691,
+      "step": 3735
+    },
+    {
+      "epoch": 5.9776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034424999999999994,
+      "loss": 2.6048,
+      "step": 3736
+    },
+    {
+      "epoch": 5.9792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034409999999999996,
+      "loss": 3.181,
+      "step": 3737
+    },
+    {
+      "epoch": 5.9808,
+      "grad_norm": 0.5898072719573975,
+      "learning_rate": 0.00034395,
+      "loss": 2.2676,
+      "step": 3738
+    },
+    {
+      "epoch": 5.9824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034379999999999995,
+      "loss": 2.8159,
+      "step": 3739
+    },
+    {
+      "epoch": 5.984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034365,
+      "loss": 2.8115,
+      "step": 3740
+    },
+    {
+      "epoch": 5.9856,
+      "grad_norm": 3.8790807723999023,
+      "learning_rate": 0.0003435,
+      "loss": 3.8424,
+      "step": 3741
+    },
+    {
+      "epoch": 5.9872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034334999999999997,
+      "loss": 3.3567,
+      "step": 3742
+    },
+    {
+      "epoch": 5.9888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034319999999999994,
+      "loss": 3.3293,
+      "step": 3743
+    },
+    {
+      "epoch": 5.9904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034304999999999996,
+      "loss": 2.1649,
+      "step": 3744
+    },
+    {
+      "epoch": 5.992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003429,
+      "loss": 2.4547,
+      "step": 3745
+    },
+    {
+      "epoch": 5.9936,
+      "grad_norm": 1.2165772914886475,
+      "learning_rate": 0.00034275,
+      "loss": 3.6393,
+      "step": 3746
+    },
+    {
+      "epoch": 5.9952,
+      "grad_norm": 0.3224068582057953,
+      "learning_rate": 0.0003425999999999999,
+      "loss": 2.719,
+      "step": 3747
+    },
+    {
+      "epoch": 5.9968,
+      "grad_norm": 0.7286518216133118,
+      "learning_rate": 0.00034244999999999995,
+      "loss": 2.3758,
+      "step": 3748
+    },
+    {
+      "epoch": 5.9984,
+      "grad_norm": 1.9613587856292725,
+      "learning_rate": 0.00034229999999999997,
+      "loss": 3.2195,
+      "step": 3749
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.1014471054077148,
+      "learning_rate": 0.00034215,
+      "loss": 2.6696,
+      "step": 3750
+    },
+    {
+      "epoch": 6.0016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034199999999999996,
+      "loss": 4.3507,
+      "step": 3751
+    },
+    {
+      "epoch": 6.0032,
+      "grad_norm": 2.0164687633514404,
+      "learning_rate": 0.00034184999999999993,
+      "loss": 3.3653,
+      "step": 3752
+    },
+    {
+      "epoch": 6.0048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034169999999999996,
+      "loss": 3.2864,
+      "step": 3753
+    },
+    {
+      "epoch": 6.0064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034155,
+      "loss": 3.4196,
+      "step": 3754
+    },
+    {
+      "epoch": 6.008,
+      "grad_norm": 0.8604856729507446,
+      "learning_rate": 0.00034139999999999995,
+      "loss": 3.6826,
+      "step": 3755
+    },
+    {
+      "epoch": 6.0096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034125,
+      "loss": 3.0471,
+      "step": 3756
+    },
+    {
+      "epoch": 6.0112,
+      "grad_norm": 4.856592178344727,
+      "learning_rate": 0.0003411,
+      "loss": 3.9458,
+      "step": 3757
+    },
+    {
+      "epoch": 6.0128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034094999999999997,
+      "loss": 3.4694,
+      "step": 3758
+    },
+    {
+      "epoch": 6.0144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034079999999999994,
+      "loss": 3.995,
+      "step": 3759
+    },
+    {
+      "epoch": 6.016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034064999999999996,
+      "loss": 3.4592,
+      "step": 3760
+    },
+    {
+      "epoch": 6.0176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003405,
+      "loss": 3.7419,
+      "step": 3761
+    },
+    {
+      "epoch": 6.0192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034035,
+      "loss": 3.5671,
+      "step": 3762
+    },
+    {
+      "epoch": 6.0208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003401999999999999,
+      "loss": 4.0311,
+      "step": 3763
+    },
+    {
+      "epoch": 6.0224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00034004999999999994,
+      "loss": 3.654,
+      "step": 3764
+    },
+    {
+      "epoch": 6.024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033989999999999997,
+      "loss": 3.0512,
+      "step": 3765
+    },
+    {
+      "epoch": 6.0256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033975,
+      "loss": 3.817,
+      "step": 3766
+    },
+    {
+      "epoch": 6.0272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033959999999999996,
+      "loss": 4.0627,
+      "step": 3767
+    },
+    {
+      "epoch": 6.0288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033944999999999993,
+      "loss": 4.4509,
+      "step": 3768
+    },
+    {
+      "epoch": 6.0304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033929999999999995,
+      "loss": 3.3457,
+      "step": 3769
+    },
+    {
+      "epoch": 6.032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033915,
+      "loss": 4.3662,
+      "step": 3770
+    },
+    {
+      "epoch": 6.0336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033899999999999995,
+      "loss": 4.1323,
+      "step": 3771
+    },
+    {
+      "epoch": 6.0352,
+      "grad_norm": 0.27863427996635437,
+      "learning_rate": 0.00033884999999999997,
+      "loss": 3.6353,
+      "step": 3772
+    },
+    {
+      "epoch": 6.0368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003387,
+      "loss": 3.9024,
+      "step": 3773
+    },
+    {
+      "epoch": 6.0384,
+      "grad_norm": 1.2410098314285278,
+      "learning_rate": 0.00033854999999999996,
+      "loss": 3.7343,
+      "step": 3774
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033839999999999993,
+      "loss": 3.529,
+      "step": 3775
+    },
+    {
+      "epoch": 6.0416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033824999999999995,
+      "loss": 4.2161,
+      "step": 3776
+    },
+    {
+      "epoch": 6.0432,
+      "grad_norm": 0.49260056018829346,
+      "learning_rate": 0.0003381,
+      "loss": 4.5646,
+      "step": 3777
+    },
+    {
+      "epoch": 6.0448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033795,
+      "loss": 3.6236,
+      "step": 3778
+    },
+    {
+      "epoch": 6.0464,
+      "grad_norm": 0.7572435140609741,
+      "learning_rate": 0.0003377999999999999,
+      "loss": 3.7258,
+      "step": 3779
+    },
+    {
+      "epoch": 6.048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033764999999999994,
+      "loss": 3.9813,
+      "step": 3780
+    },
+    {
+      "epoch": 6.0496,
+      "grad_norm": 0.10314198583364487,
+      "learning_rate": 0.00033749999999999996,
+      "loss": 3.7357,
+      "step": 3781
+    },
+    {
+      "epoch": 6.0512,
+      "grad_norm": 0.7457315921783447,
+      "learning_rate": 0.00033735,
+      "loss": 3.8844,
+      "step": 3782
+    },
+    {
+      "epoch": 6.0528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003372,
+      "loss": 4.0611,
+      "step": 3783
+    },
+    {
+      "epoch": 6.0544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033705,
+      "loss": 3.2851,
+      "step": 3784
+    },
+    {
+      "epoch": 6.056,
+      "grad_norm": 0.45380157232284546,
+      "learning_rate": 0.00033689999999999995,
+      "loss": 3.5424,
+      "step": 3785
+    },
+    {
+      "epoch": 6.0576,
+      "grad_norm": 0.24705453217029572,
+      "learning_rate": 0.00033674999999999997,
+      "loss": 3.7474,
+      "step": 3786
+    },
+    {
+      "epoch": 6.0592,
+      "grad_norm": 0.26397302746772766,
+      "learning_rate": 0.0003366,
+      "loss": 3.1184,
+      "step": 3787
+    },
+    {
+      "epoch": 6.0608,
+      "grad_norm": 0.4187479019165039,
+      "learning_rate": 0.00033644999999999997,
+      "loss": 3.7795,
+      "step": 3788
+    },
+    {
+      "epoch": 6.0624,
+      "grad_norm": 0.23197272419929504,
+      "learning_rate": 0.0003363,
+      "loss": 2.6915,
+      "step": 3789
+    },
+    {
+      "epoch": 6.064,
+      "grad_norm": 0.2890470027923584,
+      "learning_rate": 0.00033615,
+      "loss": 3.2712,
+      "step": 3790
+    },
+    {
+      "epoch": 6.0656,
+      "grad_norm": 0.8943280577659607,
+      "learning_rate": 0.000336,
+      "loss": 2.5483,
+      "step": 3791
+    },
+    {
+      "epoch": 6.0672,
+      "grad_norm": 0.8919143080711365,
+      "learning_rate": 0.00033584999999999995,
+      "loss": 2.3608,
+      "step": 3792
+    },
+    {
+      "epoch": 6.0688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003357,
+      "loss": 2.5619,
+      "step": 3793
+    },
+    {
+      "epoch": 6.0704,
+      "grad_norm": 0.5600613951683044,
+      "learning_rate": 0.00033555,
+      "loss": 3.0782,
+      "step": 3794
+    },
+    {
+      "epoch": 6.072,
+      "grad_norm": 0.7343853712081909,
+      "learning_rate": 0.0003354,
+      "loss": 3.122,
+      "step": 3795
+    },
+    {
+      "epoch": 6.0736,
+      "grad_norm": 0.7387804985046387,
+      "learning_rate": 0.00033524999999999994,
+      "loss": 2.9913,
+      "step": 3796
+    },
+    {
+      "epoch": 6.0752,
+      "grad_norm": 1.676078200340271,
+      "learning_rate": 0.00033509999999999996,
+      "loss": 2.456,
+      "step": 3797
+    },
+    {
+      "epoch": 6.0768,
+      "grad_norm": 1.1079598665237427,
+      "learning_rate": 0.00033495,
+      "loss": 2.82,
+      "step": 3798
+    },
+    {
+      "epoch": 6.0784,
+      "grad_norm": 0.6876312494277954,
+      "learning_rate": 0.0003348,
+      "loss": 2.9411,
+      "step": 3799
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": NaN,
+      "learning_rate": 0.0003348,
+      "loss": 3.124,
+      "step": 3800
+    },
+    {
+      "epoch": 6.0816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033465,
+      "loss": 4.1995,
+      "step": 3801
+    },
+    {
+      "epoch": 6.0832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033449999999999994,
+      "loss": 3.8703,
+      "step": 3802
+    },
+    {
+      "epoch": 6.0848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033434999999999997,
+      "loss": 3.5055,
+      "step": 3803
+    },
+    {
+      "epoch": 6.0864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003342,
+      "loss": 2.6898,
+      "step": 3804
+    },
+    {
+      "epoch": 6.088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033404999999999996,
+      "loss": 3.3677,
+      "step": 3805
+    },
+    {
+      "epoch": 6.0896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003339,
+      "loss": 3.135,
+      "step": 3806
+    },
+    {
+      "epoch": 6.0912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033375,
+      "loss": 3.8218,
+      "step": 3807
+    },
+    {
+      "epoch": 6.0928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003336,
+      "loss": 3.883,
+      "step": 3808
+    },
+    {
+      "epoch": 6.0944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033344999999999995,
+      "loss": 3.4605,
+      "step": 3809
+    },
+    {
+      "epoch": 6.096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033329999999999997,
+      "loss": 4.6752,
+      "step": 3810
+    },
+    {
+      "epoch": 6.0976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033315,
+      "loss": 3.5125,
+      "step": 3811
+    },
+    {
+      "epoch": 6.0992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000333,
+      "loss": 4.1268,
+      "step": 3812
+    },
+    {
+      "epoch": 6.1008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033284999999999993,
+      "loss": 3.7619,
+      "step": 3813
+    },
+    {
+      "epoch": 6.1024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033269999999999996,
+      "loss": 4.4734,
+      "step": 3814
+    },
+    {
+      "epoch": 6.104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033255,
+      "loss": 3.1928,
+      "step": 3815
+    },
+    {
+      "epoch": 6.1056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003324,
+      "loss": 3.406,
+      "step": 3816
+    },
+    {
+      "epoch": 6.1072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033224999999999997,
+      "loss": 4.1266,
+      "step": 3817
+    },
+    {
+      "epoch": 6.1088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033209999999999994,
+      "loss": 4.3521,
+      "step": 3818
+    },
+    {
+      "epoch": 6.1104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033194999999999996,
+      "loss": 3.9209,
+      "step": 3819
+    },
+    {
+      "epoch": 6.112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003318,
+      "loss": 3.5629,
+      "step": 3820
+    },
+    {
+      "epoch": 6.1136,
+      "grad_norm": 1.4520596265792847,
+      "learning_rate": 0.00033164999999999996,
+      "loss": 4.3292,
+      "step": 3821
+    },
+    {
+      "epoch": 6.1152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003315,
+      "loss": 3.9747,
+      "step": 3822
+    },
+    {
+      "epoch": 6.1168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033135,
+      "loss": 3.5212,
+      "step": 3823
+    },
+    {
+      "epoch": 6.1184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003312,
+      "loss": 4.3496,
+      "step": 3824
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033104999999999994,
+      "loss": 4.2499,
+      "step": 3825
+    },
+    {
+      "epoch": 6.1216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033089999999999997,
+      "loss": 4.6441,
+      "step": 3826
+    },
+    {
+      "epoch": 6.1232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033075,
+      "loss": 4.1746,
+      "step": 3827
+    },
+    {
+      "epoch": 6.1248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003306,
+      "loss": 4.3512,
+      "step": 3828
+    },
+    {
+      "epoch": 6.1264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033044999999999993,
+      "loss": 3.6166,
+      "step": 3829
+    },
+    {
+      "epoch": 6.128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033029999999999995,
+      "loss": 3.8687,
+      "step": 3830
+    },
+    {
+      "epoch": 6.1296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033015,
+      "loss": 3.8298,
+      "step": 3831
+    },
+    {
+      "epoch": 6.1312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00033,
+      "loss": 3.991,
+      "step": 3832
+    },
+    {
+      "epoch": 6.1328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032984999999999997,
+      "loss": 4.028,
+      "step": 3833
+    },
+    {
+      "epoch": 6.1344,
+      "grad_norm": 2.680119037628174,
+      "learning_rate": 0.0003297,
+      "loss": 3.7545,
+      "step": 3834
+    },
+    {
+      "epoch": 6.136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032954999999999996,
+      "loss": 3.6376,
+      "step": 3835
+    },
+    {
+      "epoch": 6.1376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003294,
+      "loss": 3.226,
+      "step": 3836
+    },
+    {
+      "epoch": 6.1392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032924999999999995,
+      "loss": 3.8085,
+      "step": 3837
+    },
+    {
+      "epoch": 6.1408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003291,
+      "loss": 3.4092,
+      "step": 3838
+    },
+    {
+      "epoch": 6.1424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032895,
+      "loss": 3.266,
+      "step": 3839
+    },
+    {
+      "epoch": 6.144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003288,
+      "loss": 4.4733,
+      "step": 3840
+    },
+    {
+      "epoch": 6.1456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032864999999999994,
+      "loss": 3.4369,
+      "step": 3841
+    },
+    {
+      "epoch": 6.1472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032849999999999996,
+      "loss": 3.8578,
+      "step": 3842
+    },
+    {
+      "epoch": 6.1488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032835,
+      "loss": 3.9243,
+      "step": 3843
+    },
+    {
+      "epoch": 6.1504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003282,
+      "loss": 3.7415,
+      "step": 3844
+    },
+    {
+      "epoch": 6.152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003280499999999999,
+      "loss": 2.8646,
+      "step": 3845
+    },
+    {
+      "epoch": 6.1536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032789999999999995,
+      "loss": 3.7953,
+      "step": 3846
+    },
+    {
+      "epoch": 6.1552,
+      "grad_norm": 0.2975400686264038,
+      "learning_rate": 0.00032774999999999997,
+      "loss": 3.0622,
+      "step": 3847
+    },
+    {
+      "epoch": 6.1568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003276,
+      "loss": 3.1223,
+      "step": 3848
+    },
+    {
+      "epoch": 6.1584,
+      "grad_norm": 2.694124221801758,
+      "learning_rate": 0.00032744999999999996,
+      "loss": 4.4976,
+      "step": 3849
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": NaN,
+      "learning_rate": 0.00032744999999999996,
+      "loss": 3.6087,
+      "step": 3850
+    },
+    {
+      "epoch": 6.1616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003273,
+      "loss": 7.9388,
+      "step": 3851
+    },
+    {
+      "epoch": 6.1632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032714999999999996,
+      "loss": 5.2248,
+      "step": 3852
+    },
+    {
+      "epoch": 6.1648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000327,
+      "loss": 3.9534,
+      "step": 3853
+    },
+    {
+      "epoch": 6.1664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032684999999999995,
+      "loss": 5.0484,
+      "step": 3854
+    },
+    {
+      "epoch": 6.168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032669999999999997,
+      "loss": 3.3562,
+      "step": 3855
+    },
+    {
+      "epoch": 6.1696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032655,
+      "loss": 3.0022,
+      "step": 3856
+    },
+    {
+      "epoch": 6.1712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003264,
+      "loss": 3.0776,
+      "step": 3857
+    },
+    {
+      "epoch": 6.1728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032624999999999993,
+      "loss": 4.0646,
+      "step": 3858
+    },
+    {
+      "epoch": 6.1744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032609999999999996,
+      "loss": 2.6194,
+      "step": 3859
+    },
+    {
+      "epoch": 6.176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032595,
+      "loss": 3.1227,
+      "step": 3860
+    },
+    {
+      "epoch": 6.1776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003258,
+      "loss": 3.2917,
+      "step": 3861
+    },
+    {
+      "epoch": 6.1792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003256499999999999,
+      "loss": 3.45,
+      "step": 3862
+    },
+    {
+      "epoch": 6.1808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032549999999999994,
+      "loss": 2.762,
+      "step": 3863
+    },
+    {
+      "epoch": 6.1824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032534999999999997,
+      "loss": 3.0529,
+      "step": 3864
+    },
+    {
+      "epoch": 6.184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003252,
+      "loss": 2.7619,
+      "step": 3865
+    },
+    {
+      "epoch": 6.1856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032504999999999996,
+      "loss": 3.0824,
+      "step": 3866
+    },
+    {
+      "epoch": 6.1872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003249,
+      "loss": 3.1386,
+      "step": 3867
+    },
+    {
+      "epoch": 6.1888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032474999999999995,
+      "loss": 2.5669,
+      "step": 3868
+    },
+    {
+      "epoch": 6.1904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003246,
+      "loss": 3.0091,
+      "step": 3869
+    },
+    {
+      "epoch": 6.192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032444999999999994,
+      "loss": 2.4578,
+      "step": 3870
+    },
+    {
+      "epoch": 6.1936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032429999999999997,
+      "loss": 2.9423,
+      "step": 3871
+    },
+    {
+      "epoch": 6.1952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032415,
+      "loss": 3.1918,
+      "step": 3872
+    },
+    {
+      "epoch": 6.1968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000324,
+      "loss": 2.8123,
+      "step": 3873
+    },
+    {
+      "epoch": 6.1984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032384999999999993,
+      "loss": 3.1704,
+      "step": 3874
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032369999999999995,
+      "loss": 2.6135,
+      "step": 3875
+    },
+    {
+      "epoch": 6.2016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032355,
+      "loss": 3.4235,
+      "step": 3876
+    },
+    {
+      "epoch": 6.2032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003234,
+      "loss": 2.5807,
+      "step": 3877
+    },
+    {
+      "epoch": 6.2048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003232499999999999,
+      "loss": 2.2725,
+      "step": 3878
+    },
+    {
+      "epoch": 6.2064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032309999999999994,
+      "loss": 3.3268,
+      "step": 3879
+    },
+    {
+      "epoch": 6.208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032294999999999996,
+      "loss": 2.7539,
+      "step": 3880
+    },
+    {
+      "epoch": 6.2096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003228,
+      "loss": 2.1264,
+      "step": 3881
+    },
+    {
+      "epoch": 6.2112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032264999999999996,
+      "loss": 2.7397,
+      "step": 3882
+    },
+    {
+      "epoch": 6.2128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003225,
+      "loss": 2.8085,
+      "step": 3883
+    },
+    {
+      "epoch": 6.2144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032235,
+      "loss": 2.5252,
+      "step": 3884
+    },
+    {
+      "epoch": 6.216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032219999999999997,
+      "loss": 2.4381,
+      "step": 3885
+    },
+    {
+      "epoch": 6.2176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032204999999999994,
+      "loss": 2.9784,
+      "step": 3886
+    },
+    {
+      "epoch": 6.2192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032189999999999996,
+      "loss": 2.8678,
+      "step": 3887
+    },
+    {
+      "epoch": 6.2208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032175,
+      "loss": 2.3483,
+      "step": 3888
+    },
+    {
+      "epoch": 6.2224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003216,
+      "loss": 2.6857,
+      "step": 3889
+    },
+    {
+      "epoch": 6.224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003214499999999999,
+      "loss": 3.1904,
+      "step": 3890
+    },
+    {
+      "epoch": 6.2256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032129999999999995,
+      "loss": 4.1369,
+      "step": 3891
+    },
+    {
+      "epoch": 6.2272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032114999999999997,
+      "loss": 3.0205,
+      "step": 3892
+    },
+    {
+      "epoch": 6.2288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000321,
+      "loss": 2.5759,
+      "step": 3893
+    },
+    {
+      "epoch": 6.2304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032084999999999997,
+      "loss": 2.9976,
+      "step": 3894
+    },
+    {
+      "epoch": 6.232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032069999999999993,
+      "loss": 3.7825,
+      "step": 3895
+    },
+    {
+      "epoch": 6.2336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032054999999999996,
+      "loss": 3.623,
+      "step": 3896
+    },
+    {
+      "epoch": 6.2352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003204,
+      "loss": 3.3718,
+      "step": 3897
+    },
+    {
+      "epoch": 6.2368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00032024999999999995,
+      "loss": 2.7264,
+      "step": 3898
+    },
+    {
+      "epoch": 6.2384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003201,
+      "loss": 4.1795,
+      "step": 3899
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031995,
+      "loss": 3.4573,
+      "step": 3900
+    },
+    {
+      "epoch": 6.2416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031979999999999997,
+      "loss": 3.8604,
+      "step": 3901
+    },
+    {
+      "epoch": 6.2432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031964999999999994,
+      "loss": 3.1814,
+      "step": 3902
+    },
+    {
+      "epoch": 6.2448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031949999999999996,
+      "loss": 3.407,
+      "step": 3903
+    },
+    {
+      "epoch": 6.2464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031935,
+      "loss": 3.8178,
+      "step": 3904
+    },
+    {
+      "epoch": 6.248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003192,
+      "loss": 3.2901,
+      "step": 3905
+    },
+    {
+      "epoch": 6.2496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003190499999999999,
+      "loss": 5.0435,
+      "step": 3906
+    },
+    {
+      "epoch": 6.2512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031889999999999995,
+      "loss": 3.0831,
+      "step": 3907
+    },
+    {
+      "epoch": 6.2528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031874999999999997,
+      "loss": 3.5521,
+      "step": 3908
+    },
+    {
+      "epoch": 6.2544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003186,
+      "loss": 2.5273,
+      "step": 3909
+    },
+    {
+      "epoch": 6.256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031845,
+      "loss": 2.1646,
+      "step": 3910
+    },
+    {
+      "epoch": 6.2576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031829999999999993,
+      "loss": 2.8208,
+      "step": 3911
+    },
+    {
+      "epoch": 6.2592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031814999999999995,
+      "loss": 3.3235,
+      "step": 3912
+    },
+    {
+      "epoch": 6.2608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000318,
+      "loss": 2.6847,
+      "step": 3913
+    },
+    {
+      "epoch": 6.2624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031785,
+      "loss": 2.4997,
+      "step": 3914
+    },
+    {
+      "epoch": 6.264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031769999999999997,
+      "loss": 2.3451,
+      "step": 3915
+    },
+    {
+      "epoch": 6.2656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031755,
+      "loss": 3.2673,
+      "step": 3916
+    },
+    {
+      "epoch": 6.2672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031739999999999996,
+      "loss": 2.8985,
+      "step": 3917
+    },
+    {
+      "epoch": 6.2688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031725,
+      "loss": 2.4833,
+      "step": 3918
+    },
+    {
+      "epoch": 6.2704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031709999999999996,
+      "loss": 2.8448,
+      "step": 3919
+    },
+    {
+      "epoch": 6.272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031695,
+      "loss": 2.5166,
+      "step": 3920
+    },
+    {
+      "epoch": 6.2736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003168,
+      "loss": 3.3755,
+      "step": 3921
+    },
+    {
+      "epoch": 6.2752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031665,
+      "loss": 2.3369,
+      "step": 3922
+    },
+    {
+      "epoch": 6.2768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031649999999999994,
+      "loss": 2.9503,
+      "step": 3923
+    },
+    {
+      "epoch": 6.2783999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031634999999999996,
+      "loss": 2.8635,
+      "step": 3924
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003162,
+      "loss": 2.9627,
+      "step": 3925
+    },
+    {
+      "epoch": 6.2816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031605,
+      "loss": 2.7479,
+      "step": 3926
+    },
+    {
+      "epoch": 6.2832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003158999999999999,
+      "loss": 2.2887,
+      "step": 3927
+    },
+    {
+      "epoch": 6.2848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031574999999999995,
+      "loss": 3.7974,
+      "step": 3928
+    },
+    {
+      "epoch": 6.2864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003156,
+      "loss": 3.4341,
+      "step": 3929
+    },
+    {
+      "epoch": 6.288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031545,
+      "loss": 2.3601,
+      "step": 3930
+    },
+    {
+      "epoch": 6.2896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031529999999999997,
+      "loss": 3.3701,
+      "step": 3931
+    },
+    {
+      "epoch": 6.2912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031515,
+      "loss": 3.1594,
+      "step": 3932
+    },
+    {
+      "epoch": 6.2928,
+      "grad_norm": 0.7066333889961243,
+      "learning_rate": 0.00031499999999999996,
+      "loss": 2.9424,
+      "step": 3933
+    },
+    {
+      "epoch": 6.2943999999999996,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031485,
+      "loss": 2.5434,
+      "step": 3934
+    },
+    {
+      "epoch": 6.296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031469999999999995,
+      "loss": 2.9575,
+      "step": 3935
+    },
+    {
+      "epoch": 6.2976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031455,
+      "loss": 2.647,
+      "step": 3936
+    },
+    {
+      "epoch": 6.2992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003144,
+      "loss": 2.5617,
+      "step": 3937
+    },
+    {
+      "epoch": 6.3008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031425,
+      "loss": 4.2255,
+      "step": 3938
+    },
+    {
+      "epoch": 6.3024000000000004,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031409999999999994,
+      "loss": 4.08,
+      "step": 3939
+    },
+    {
+      "epoch": 6.304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031394999999999996,
+      "loss": 2.8675,
+      "step": 3940
+    },
+    {
+      "epoch": 6.3056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003138,
+      "loss": 2.7682,
+      "step": 3941
+    },
+    {
+      "epoch": 6.3072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031365,
+      "loss": 2.8993,
+      "step": 3942
+    },
+    {
+      "epoch": 6.3088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003135,
+      "loss": 3.5349,
+      "step": 3943
+    },
+    {
+      "epoch": 6.3104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031334999999999995,
+      "loss": 3.5218,
+      "step": 3944
+    },
+    {
+      "epoch": 6.312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031319999999999997,
+      "loss": 3.2293,
+      "step": 3945
+    },
+    {
+      "epoch": 6.3136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031305,
+      "loss": 2.6338,
+      "step": 3946
+    },
+    {
+      "epoch": 6.3152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031289999999999996,
+      "loss": 3.2263,
+      "step": 3947
+    },
+    {
+      "epoch": 6.3168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031275,
+      "loss": 2.486,
+      "step": 3948
+    },
+    {
+      "epoch": 6.3184000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003126,
+      "loss": 3.832,
+      "step": 3949
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031245,
+      "loss": 3.0282,
+      "step": 3950
+    },
+    {
+      "epoch": 6.3216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031229999999999995,
+      "loss": 4.0251,
+      "step": 3951
+    },
+    {
+      "epoch": 6.3232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031214999999999997,
+      "loss": 4.8373,
+      "step": 3952
+    },
+    {
+      "epoch": 6.3248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000312,
+      "loss": 3.9881,
+      "step": 3953
+    },
+    {
+      "epoch": 6.3264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031185,
+      "loss": 3.7715,
+      "step": 3954
+    },
+    {
+      "epoch": 6.328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031169999999999993,
+      "loss": 3.0073,
+      "step": 3955
+    },
+    {
+      "epoch": 6.3296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031154999999999996,
+      "loss": 3.559,
+      "step": 3956
+    },
+    {
+      "epoch": 6.3312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003114,
+      "loss": 3.7248,
+      "step": 3957
+    },
+    {
+      "epoch": 6.3328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031125,
+      "loss": 3.5672,
+      "step": 3958
+    },
+    {
+      "epoch": 6.3344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031109999999999997,
+      "loss": 2.905,
+      "step": 3959
+    },
+    {
+      "epoch": 6.336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031094999999999994,
+      "loss": 3.4616,
+      "step": 3960
+    },
+    {
+      "epoch": 6.3376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031079999999999997,
+      "loss": 3.2915,
+      "step": 3961
+    },
+    {
+      "epoch": 6.3392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031065,
+      "loss": 3.3367,
+      "step": 3962
+    },
+    {
+      "epoch": 6.3408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031049999999999996,
+      "loss": 2.7731,
+      "step": 3963
+    },
+    {
+      "epoch": 6.3424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031035,
+      "loss": 2.9709,
+      "step": 3964
+    },
+    {
+      "epoch": 6.344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003102,
+      "loss": 3.2344,
+      "step": 3965
+    },
+    {
+      "epoch": 6.3456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00031005,
+      "loss": 3.5747,
+      "step": 3966
+    },
+    {
+      "epoch": 6.3472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030989999999999994,
+      "loss": 3.3665,
+      "step": 3967
+    },
+    {
+      "epoch": 6.3488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030974999999999997,
+      "loss": 3.1321,
+      "step": 3968
+    },
+    {
+      "epoch": 6.3504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003096,
+      "loss": 2.6795,
+      "step": 3969
+    },
+    {
+      "epoch": 6.352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030945,
+      "loss": 3.3899,
+      "step": 3970
+    },
+    {
+      "epoch": 6.3536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030929999999999993,
+      "loss": 2.9092,
+      "step": 3971
+    },
+    {
+      "epoch": 6.3552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030914999999999995,
+      "loss": 4.6425,
+      "step": 3972
+    },
+    {
+      "epoch": 6.3568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000309,
+      "loss": 3.0781,
+      "step": 3973
+    },
+    {
+      "epoch": 6.3584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030885,
+      "loss": 2.8645,
+      "step": 3974
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030869999999999997,
+      "loss": 2.3022,
+      "step": 3975
+    },
+    {
+      "epoch": 6.3616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030854999999999994,
+      "loss": 2.7279,
+      "step": 3976
+    },
+    {
+      "epoch": 6.3632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030839999999999996,
+      "loss": 2.7442,
+      "step": 3977
+    },
+    {
+      "epoch": 6.3648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030825,
+      "loss": 3.519,
+      "step": 3978
+    },
+    {
+      "epoch": 6.3664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030809999999999995,
+      "loss": 2.9618,
+      "step": 3979
+    },
+    {
+      "epoch": 6.368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030795,
+      "loss": 3.3267,
+      "step": 3980
+    },
+    {
+      "epoch": 6.3696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003078,
+      "loss": 2.9491,
+      "step": 3981
+    },
+    {
+      "epoch": 6.3712,
+      "grad_norm": 3.947984457015991,
+      "learning_rate": 0.00030764999999999997,
+      "loss": 3.9896,
+      "step": 3982
+    },
+    {
+      "epoch": 6.3728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030749999999999994,
+      "loss": 3.7426,
+      "step": 3983
+    },
+    {
+      "epoch": 6.3744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030734999999999996,
+      "loss": 2.8783,
+      "step": 3984
+    },
+    {
+      "epoch": 6.376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003072,
+      "loss": 2.4963,
+      "step": 3985
+    },
+    {
+      "epoch": 6.3776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030705,
+      "loss": 3.3298,
+      "step": 3986
+    },
+    {
+      "epoch": 6.3792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003068999999999999,
+      "loss": 3.0442,
+      "step": 3987
+    },
+    {
+      "epoch": 6.3808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030674999999999995,
+      "loss": 2.7951,
+      "step": 3988
+    },
+    {
+      "epoch": 6.3824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030659999999999997,
+      "loss": 3.2513,
+      "step": 3989
+    },
+    {
+      "epoch": 6.384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030645,
+      "loss": 3.1854,
+      "step": 3990
+    },
+    {
+      "epoch": 6.3856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030629999999999996,
+      "loss": 2.8245,
+      "step": 3991
+    },
+    {
+      "epoch": 6.3872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030615,
+      "loss": 3.0166,
+      "step": 3992
+    },
+    {
+      "epoch": 6.3888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030599999999999996,
+      "loss": 3.1391,
+      "step": 3993
+    },
+    {
+      "epoch": 6.3904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030585,
+      "loss": 3.8282,
+      "step": 3994
+    },
+    {
+      "epoch": 6.392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030569999999999995,
+      "loss": 2.4871,
+      "step": 3995
+    },
+    {
+      "epoch": 6.3936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030555,
+      "loss": 3.2249,
+      "step": 3996
+    },
+    {
+      "epoch": 6.3952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003054,
+      "loss": 3.1221,
+      "step": 3997
+    },
+    {
+      "epoch": 6.3968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030525,
+      "loss": 3.1722,
+      "step": 3998
+    },
+    {
+      "epoch": 6.3984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030509999999999994,
+      "loss": 3.5031,
+      "step": 3999
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": NaN,
+      "learning_rate": 0.00030509999999999994,
+      "loss": 3.3134,
+      "step": 4000
+    },
+    {
+      "epoch": 6.4,
+      "eval_cer": 0.5204408657870611,
+      "eval_loss": 3.302504777908325,
+      "eval_runtime": 159.6303,
+      "eval_samples_per_second": 19.645,
+      "eval_steps_per_second": 1.228,
+      "eval_wer": 0.8032742720961559,
+      "step": 4000
+    },
+    {
+      "epoch": 6.4016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030494999999999996,
+      "loss": 3.7485,
+      "step": 4001
+    },
+    {
+      "epoch": 6.4032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003048,
+      "loss": 3.6302,
+      "step": 4002
+    },
+    {
+      "epoch": 6.4048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030465,
+      "loss": 3.312,
+      "step": 4003
+    },
+    {
+      "epoch": 6.4064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003044999999999999,
+      "loss": 4.4313,
+      "step": 4004
+    },
+    {
+      "epoch": 6.408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030434999999999994,
+      "loss": 3.9922,
+      "step": 4005
+    },
+    {
+      "epoch": 6.4096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030419999999999997,
+      "loss": 3.5101,
+      "step": 4006
+    },
+    {
+      "epoch": 6.4112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030405,
+      "loss": 3.2452,
+      "step": 4007
+    },
+    {
+      "epoch": 6.4128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030389999999999996,
+      "loss": 3.2645,
+      "step": 4008
+    },
+    {
+      "epoch": 6.4144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030375,
+      "loss": 3.4592,
+      "step": 4009
+    },
+    {
+      "epoch": 6.416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030359999999999995,
+      "loss": 4.1544,
+      "step": 4010
+    },
+    {
+      "epoch": 6.4176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030345,
+      "loss": 3.4224,
+      "step": 4011
+    },
+    {
+      "epoch": 6.4192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030329999999999995,
+      "loss": 2.963,
+      "step": 4012
+    },
+    {
+      "epoch": 6.4208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030314999999999997,
+      "loss": 3.6251,
+      "step": 4013
+    },
+    {
+      "epoch": 6.4224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000303,
+      "loss": 2.4698,
+      "step": 4014
+    },
+    {
+      "epoch": 6.424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030285,
+      "loss": 2.7059,
+      "step": 4015
+    },
+    {
+      "epoch": 6.4256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030269999999999993,
+      "loss": 3.056,
+      "step": 4016
+    },
+    {
+      "epoch": 6.4272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030254999999999995,
+      "loss": 3.0482,
+      "step": 4017
+    },
+    {
+      "epoch": 6.4288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003024,
+      "loss": 2.4599,
+      "step": 4018
+    },
+    {
+      "epoch": 6.4304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030225,
+      "loss": 2.67,
+      "step": 4019
+    },
+    {
+      "epoch": 6.432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003020999999999999,
+      "loss": 3.273,
+      "step": 4020
+    },
+    {
+      "epoch": 6.4336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030194999999999994,
+      "loss": 2.584,
+      "step": 4021
+    },
+    {
+      "epoch": 6.4352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030179999999999996,
+      "loss": 2.8512,
+      "step": 4022
+    },
+    {
+      "epoch": 6.4368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030165,
+      "loss": 3.4812,
+      "step": 4023
+    },
+    {
+      "epoch": 6.4384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030149999999999996,
+      "loss": 3.0326,
+      "step": 4024
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030135,
+      "loss": 2.572,
+      "step": 4025
+    },
+    {
+      "epoch": 6.4416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030119999999999995,
+      "loss": 2.8446,
+      "step": 4026
+    },
+    {
+      "epoch": 6.4432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030104999999999997,
+      "loss": 3.3667,
+      "step": 4027
+    },
+    {
+      "epoch": 6.4448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030089999999999994,
+      "loss": 2.9241,
+      "step": 4028
+    },
+    {
+      "epoch": 6.4464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030074999999999996,
+      "loss": 3.1371,
+      "step": 4029
+    },
+    {
+      "epoch": 6.448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003006,
+      "loss": 2.3366,
+      "step": 4030
+    },
+    {
+      "epoch": 6.4496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030045,
+      "loss": 3.2662,
+      "step": 4031
+    },
+    {
+      "epoch": 6.4512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003002999999999999,
+      "loss": 2.4098,
+      "step": 4032
+    },
+    {
+      "epoch": 6.4528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00030014999999999995,
+      "loss": 3.1829,
+      "step": 4033
+    },
+    {
+      "epoch": 6.4544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0003,
+      "loss": 3.1601,
+      "step": 4034
+    },
+    {
+      "epoch": 6.456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029985,
+      "loss": 3.1077,
+      "step": 4035
+    },
+    {
+      "epoch": 6.4576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029969999999999997,
+      "loss": 2.7685,
+      "step": 4036
+    },
+    {
+      "epoch": 6.4592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029955,
+      "loss": 2.6426,
+      "step": 4037
+    },
+    {
+      "epoch": 6.4608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029939999999999996,
+      "loss": 3.2698,
+      "step": 4038
+    },
+    {
+      "epoch": 6.4624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029925,
+      "loss": 3.1552,
+      "step": 4039
+    },
+    {
+      "epoch": 6.464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029909999999999995,
+      "loss": 3.6805,
+      "step": 4040
+    },
+    {
+      "epoch": 6.4656,
+      "grad_norm": NaN,
+      "learning_rate": 0.00029909999999999995,
+      "loss": 1.2008,
+      "step": 4041
+    },
+    {
+      "epoch": 6.4672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029895,
+      "loss": 2.9159,
+      "step": 4042
+    },
+    {
+      "epoch": 6.4688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002988,
+      "loss": 3.4111,
+      "step": 4043
+    },
+    {
+      "epoch": 6.4704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029864999999999997,
+      "loss": 2.6915,
+      "step": 4044
+    },
+    {
+      "epoch": 6.4719999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002985,
+      "loss": 2.7053,
+      "step": 4045
+    },
+    {
+      "epoch": 6.4736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029835,
+      "loss": 2.8908,
+      "step": 4046
+    },
+    {
+      "epoch": 6.4752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002982,
+      "loss": 3.1514,
+      "step": 4047
+    },
+    {
+      "epoch": 6.4768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029805,
+      "loss": 2.9263,
+      "step": 4048
+    },
+    {
+      "epoch": 6.4784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002979,
+      "loss": 3.1634,
+      "step": 4049
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029775,
+      "loss": 3.4911,
+      "step": 4050
+    },
+    {
+      "epoch": 6.4816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029759999999999997,
+      "loss": 4.6418,
+      "step": 4051
+    },
+    {
+      "epoch": 6.4832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029745,
+      "loss": 4.2075,
+      "step": 4052
+    },
+    {
+      "epoch": 6.4848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029729999999999996,
+      "loss": 3.1241,
+      "step": 4053
+    },
+    {
+      "epoch": 6.4864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029715,
+      "loss": 3.7562,
+      "step": 4054
+    },
+    {
+      "epoch": 6.4879999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 3.0509,
+      "step": 4055
+    },
+    {
+      "epoch": 6.4896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029685,
+      "loss": 3.2858,
+      "step": 4056
+    },
+    {
+      "epoch": 6.4912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029669999999999995,
+      "loss": 3.0402,
+      "step": 4057
+    },
+    {
+      "epoch": 6.4928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029654999999999997,
+      "loss": 3.7944,
+      "step": 4058
+    },
+    {
+      "epoch": 6.4944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002964,
+      "loss": 2.8896,
+      "step": 4059
+    },
+    {
+      "epoch": 6.496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029624999999999996,
+      "loss": 2.9788,
+      "step": 4060
+    },
+    {
+      "epoch": 6.4976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002961,
+      "loss": 3.1403,
+      "step": 4061
+    },
+    {
+      "epoch": 6.4992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029595,
+      "loss": 3.5309,
+      "step": 4062
+    },
+    {
+      "epoch": 6.5008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002958,
+      "loss": 2.9013,
+      "step": 4063
+    },
+    {
+      "epoch": 6.5024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029565,
+      "loss": 3.0639,
+      "step": 4064
+    },
+    {
+      "epoch": 6.504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029549999999999997,
+      "loss": 3.1012,
+      "step": 4065
+    },
+    {
+      "epoch": 6.5056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029535,
+      "loss": 2.8293,
+      "step": 4066
+    },
+    {
+      "epoch": 6.5072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029519999999999997,
+      "loss": 2.6492,
+      "step": 4067
+    },
+    {
+      "epoch": 6.5088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029505,
+      "loss": 2.7127,
+      "step": 4068
+    },
+    {
+      "epoch": 6.5104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029489999999999996,
+      "loss": 2.9354,
+      "step": 4069
+    },
+    {
+      "epoch": 6.5120000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029475,
+      "loss": 3.0551,
+      "step": 4070
+    },
+    {
+      "epoch": 6.5136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029459999999999995,
+      "loss": 2.4176,
+      "step": 4071
+    },
+    {
+      "epoch": 6.5152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029445,
+      "loss": 2.4007,
+      "step": 4072
+    },
+    {
+      "epoch": 6.5168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029429999999999994,
+      "loss": 2.4701,
+      "step": 4073
+    },
+    {
+      "epoch": 6.5184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029414999999999997,
+      "loss": 2.8651,
+      "step": 4074
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000294,
+      "loss": 2.2248,
+      "step": 4075
+    },
+    {
+      "epoch": 6.5216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029384999999999996,
+      "loss": 2.7142,
+      "step": 4076
+    },
+    {
+      "epoch": 6.5232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002937,
+      "loss": 3.0723,
+      "step": 4077
+    },
+    {
+      "epoch": 6.5248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029355,
+      "loss": 2.6237,
+      "step": 4078
+    },
+    {
+      "epoch": 6.5264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002934,
+      "loss": 2.7589,
+      "step": 4079
+    },
+    {
+      "epoch": 6.5280000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029325,
+      "loss": 2.5206,
+      "step": 4080
+    },
+    {
+      "epoch": 6.5296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029309999999999997,
+      "loss": 2.2624,
+      "step": 4081
+    },
+    {
+      "epoch": 6.5312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029295,
+      "loss": 2.7337,
+      "step": 4082
+    },
+    {
+      "epoch": 6.5328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029279999999999996,
+      "loss": 3.2515,
+      "step": 4083
+    },
+    {
+      "epoch": 6.5344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029265,
+      "loss": 3.1189,
+      "step": 4084
+    },
+    {
+      "epoch": 6.536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029249999999999995,
+      "loss": 3.1952,
+      "step": 4085
+    },
+    {
+      "epoch": 6.5376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029235,
+      "loss": 3.007,
+      "step": 4086
+    },
+    {
+      "epoch": 6.5392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029219999999999995,
+      "loss": 2.7831,
+      "step": 4087
+    },
+    {
+      "epoch": 6.5408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029204999999999997,
+      "loss": 2.8143,
+      "step": 4088
+    },
+    {
+      "epoch": 6.5424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002919,
+      "loss": 2.461,
+      "step": 4089
+    },
+    {
+      "epoch": 6.5440000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029174999999999996,
+      "loss": 2.4208,
+      "step": 4090
+    },
+    {
+      "epoch": 6.5456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002916,
+      "loss": 2.7209,
+      "step": 4091
+    },
+    {
+      "epoch": 6.5472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029145,
+      "loss": 2.8859,
+      "step": 4092
+    },
+    {
+      "epoch": 6.5488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002913,
+      "loss": 2.6131,
+      "step": 4093
+    },
+    {
+      "epoch": 6.5504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029115,
+      "loss": 2.8062,
+      "step": 4094
+    },
+    {
+      "epoch": 6.552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029099999999999997,
+      "loss": 2.6563,
+      "step": 4095
+    },
+    {
+      "epoch": 6.5536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029085,
+      "loss": 2.732,
+      "step": 4096
+    },
+    {
+      "epoch": 6.5552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029069999999999996,
+      "loss": 3.2932,
+      "step": 4097
+    },
+    {
+      "epoch": 6.5568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029055,
+      "loss": 3.189,
+      "step": 4098
+    },
+    {
+      "epoch": 6.5584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029039999999999996,
+      "loss": 3.0375,
+      "step": 4099
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029025,
+      "loss": 4.1397,
+      "step": 4100
+    },
+    {
+      "epoch": 6.5616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00029009999999999995,
+      "loss": 4.448,
+      "step": 4101
+    },
+    {
+      "epoch": 6.5632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028995,
+      "loss": 3.241,
+      "step": 4102
+    },
+    {
+      "epoch": 6.5648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028979999999999994,
+      "loss": 3.3749,
+      "step": 4103
+    },
+    {
+      "epoch": 6.5664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028964999999999997,
+      "loss": 3.301,
+      "step": 4104
+    },
+    {
+      "epoch": 6.568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002895,
+      "loss": 3.2784,
+      "step": 4105
+    },
+    {
+      "epoch": 6.5696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028934999999999996,
+      "loss": 3.8074,
+      "step": 4106
+    },
+    {
+      "epoch": 6.5712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002892,
+      "loss": 3.412,
+      "step": 4107
+    },
+    {
+      "epoch": 6.5728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028905,
+      "loss": 2.9879,
+      "step": 4108
+    },
+    {
+      "epoch": 6.5744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002889,
+      "loss": 4.0027,
+      "step": 4109
+    },
+    {
+      "epoch": 6.576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028875,
+      "loss": 2.685,
+      "step": 4110
+    },
+    {
+      "epoch": 6.5776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028859999999999997,
+      "loss": 2.8833,
+      "step": 4111
+    },
+    {
+      "epoch": 6.5792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028845,
+      "loss": 2.7789,
+      "step": 4112
+    },
+    {
+      "epoch": 6.5808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028829999999999996,
+      "loss": 3.5811,
+      "step": 4113
+    },
+    {
+      "epoch": 6.5824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028815,
+      "loss": 2.6637,
+      "step": 4114
+    },
+    {
+      "epoch": 6.584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 2.8443,
+      "step": 4115
+    },
+    {
+      "epoch": 6.5856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028785,
+      "loss": 3.1588,
+      "step": 4116
+    },
+    {
+      "epoch": 6.5872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028769999999999995,
+      "loss": 2.6272,
+      "step": 4117
+    },
+    {
+      "epoch": 6.5888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028754999999999997,
+      "loss": 3.33,
+      "step": 4118
+    },
+    {
+      "epoch": 6.5904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028739999999999994,
+      "loss": 2.8541,
+      "step": 4119
+    },
+    {
+      "epoch": 6.592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028724999999999996,
+      "loss": 2.3283,
+      "step": 4120
+    },
+    {
+      "epoch": 6.5936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002871,
+      "loss": 2.9972,
+      "step": 4121
+    },
+    {
+      "epoch": 6.5952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028694999999999995,
+      "loss": 3.2081,
+      "step": 4122
+    },
+    {
+      "epoch": 6.5968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002868,
+      "loss": 3.4921,
+      "step": 4123
+    },
+    {
+      "epoch": 6.5984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028665,
+      "loss": 2.5474,
+      "step": 4124
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028649999999999997,
+      "loss": 3.0886,
+      "step": 4125
+    },
+    {
+      "epoch": 6.6016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028635,
+      "loss": 2.9268,
+      "step": 4126
+    },
+    {
+      "epoch": 6.6032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028619999999999996,
+      "loss": 2.3568,
+      "step": 4127
+    },
+    {
+      "epoch": 6.6048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028605,
+      "loss": 3.1724,
+      "step": 4128
+    },
+    {
+      "epoch": 6.6064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028589999999999996,
+      "loss": 2.4307,
+      "step": 4129
+    },
+    {
+      "epoch": 6.608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028575,
+      "loss": 2.8321,
+      "step": 4130
+    },
+    {
+      "epoch": 6.6096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028559999999999995,
+      "loss": 2.5707,
+      "step": 4131
+    },
+    {
+      "epoch": 6.6112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028544999999999997,
+      "loss": 3.0704,
+      "step": 4132
+    },
+    {
+      "epoch": 6.6128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028529999999999994,
+      "loss": 3.2784,
+      "step": 4133
+    },
+    {
+      "epoch": 6.6144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028514999999999997,
+      "loss": 3.3985,
+      "step": 4134
+    },
+    {
+      "epoch": 6.616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000285,
+      "loss": 2.8685,
+      "step": 4135
+    },
+    {
+      "epoch": 6.6176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028484999999999996,
+      "loss": 3.0241,
+      "step": 4136
+    },
+    {
+      "epoch": 6.6192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002847,
+      "loss": 2.4871,
+      "step": 4137
+    },
+    {
+      "epoch": 6.6208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028455,
+      "loss": 2.8786,
+      "step": 4138
+    },
+    {
+      "epoch": 6.6224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002844,
+      "loss": 2.8877,
+      "step": 4139
+    },
+    {
+      "epoch": 6.624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028425,
+      "loss": 2.6461,
+      "step": 4140
+    },
+    {
+      "epoch": 6.6256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028409999999999997,
+      "loss": 2.699,
+      "step": 4141
+    },
+    {
+      "epoch": 6.6272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028395,
+      "loss": 2.8822,
+      "step": 4142
+    },
+    {
+      "epoch": 6.6288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028379999999999996,
+      "loss": 2.7284,
+      "step": 4143
+    },
+    {
+      "epoch": 6.6304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028365,
+      "loss": 2.8397,
+      "step": 4144
+    },
+    {
+      "epoch": 6.632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028349999999999995,
+      "loss": 2.984,
+      "step": 4145
+    },
+    {
+      "epoch": 6.6336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028335,
+      "loss": 3.0078,
+      "step": 4146
+    },
+    {
+      "epoch": 6.6352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028319999999999994,
+      "loss": 3.1687,
+      "step": 4147
+    },
+    {
+      "epoch": 6.6368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028304999999999997,
+      "loss": 3.0273,
+      "step": 4148
+    },
+    {
+      "epoch": 6.6384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028289999999999994,
+      "loss": 3.1394,
+      "step": 4149
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028274999999999996,
+      "loss": 4.6158,
+      "step": 4150
+    },
+    {
+      "epoch": 6.6416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002826,
+      "loss": 3.8414,
+      "step": 4151
+    },
+    {
+      "epoch": 6.6432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028244999999999995,
+      "loss": 2.765,
+      "step": 4152
+    },
+    {
+      "epoch": 6.6448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002823,
+      "loss": 3.6998,
+      "step": 4153
+    },
+    {
+      "epoch": 6.6464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028215,
+      "loss": 5.3172,
+      "step": 4154
+    },
+    {
+      "epoch": 6.648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 2.9695,
+      "step": 4155
+    },
+    {
+      "epoch": 6.6495999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028185,
+      "loss": 3.8436,
+      "step": 4156
+    },
+    {
+      "epoch": 6.6512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028169999999999996,
+      "loss": 3.3066,
+      "step": 4157
+    },
+    {
+      "epoch": 6.6528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028155,
+      "loss": 3.0264,
+      "step": 4158
+    },
+    {
+      "epoch": 6.6544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028139999999999996,
+      "loss": 2.5204,
+      "step": 4159
+    },
+    {
+      "epoch": 6.656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028125,
+      "loss": 2.4504,
+      "step": 4160
+    },
+    {
+      "epoch": 6.6576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002811,
+      "loss": 3.045,
+      "step": 4161
+    },
+    {
+      "epoch": 6.6592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028094999999999997,
+      "loss": 3.8107,
+      "step": 4162
+    },
+    {
+      "epoch": 6.6608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002808,
+      "loss": 2.8968,
+      "step": 4163
+    },
+    {
+      "epoch": 6.6624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028064999999999996,
+      "loss": 2.5925,
+      "step": 4164
+    },
+    {
+      "epoch": 6.664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002805,
+      "loss": 3.1626,
+      "step": 4165
+    },
+    {
+      "epoch": 6.6655999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028034999999999996,
+      "loss": 3.3632,
+      "step": 4166
+    },
+    {
+      "epoch": 6.6672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002802,
+      "loss": 2.757,
+      "step": 4167
+    },
+    {
+      "epoch": 6.6688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00028004999999999995,
+      "loss": 2.8508,
+      "step": 4168
+    },
+    {
+      "epoch": 6.6704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002799,
+      "loss": 2.8919,
+      "step": 4169
+    },
+    {
+      "epoch": 6.672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027975,
+      "loss": 2.7713,
+      "step": 4170
+    },
+    {
+      "epoch": 6.6736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027959999999999997,
+      "loss": 3.0,
+      "step": 4171
+    },
+    {
+      "epoch": 6.6752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027945,
+      "loss": 2.4242,
+      "step": 4172
+    },
+    {
+      "epoch": 6.6768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002793,
+      "loss": 2.288,
+      "step": 4173
+    },
+    {
+      "epoch": 6.6784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027915,
+      "loss": 2.9365,
+      "step": 4174
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000279,
+      "loss": 2.5699,
+      "step": 4175
+    },
+    {
+      "epoch": 6.6815999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027885,
+      "loss": 2.9192,
+      "step": 4176
+    },
+    {
+      "epoch": 6.6832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002787,
+      "loss": 2.8393,
+      "step": 4177
+    },
+    {
+      "epoch": 6.6848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027854999999999997,
+      "loss": 2.9024,
+      "step": 4178
+    },
+    {
+      "epoch": 6.6864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002784,
+      "loss": 2.3331,
+      "step": 4179
+    },
+    {
+      "epoch": 6.688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027824999999999996,
+      "loss": 2.8534,
+      "step": 4180
+    },
+    {
+      "epoch": 6.6896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002781,
+      "loss": 2.1188,
+      "step": 4181
+    },
+    {
+      "epoch": 6.6912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027794999999999995,
+      "loss": 2.5233,
+      "step": 4182
+    },
+    {
+      "epoch": 6.6928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002778,
+      "loss": 2.1454,
+      "step": 4183
+    },
+    {
+      "epoch": 6.6944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027764999999999995,
+      "loss": 2.5926,
+      "step": 4184
+    },
+    {
+      "epoch": 6.696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027749999999999997,
+      "loss": 2.6491,
+      "step": 4185
+    },
+    {
+      "epoch": 6.6975999999999996,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027735,
+      "loss": 3.3521,
+      "step": 4186
+    },
+    {
+      "epoch": 6.6992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002772,
+      "loss": 2.8787,
+      "step": 4187
+    },
+    {
+      "epoch": 6.7008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027705,
+      "loss": 2.3401,
+      "step": 4188
+    },
+    {
+      "epoch": 6.7024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002769,
+      "loss": 2.6153,
+      "step": 4189
+    },
+    {
+      "epoch": 6.704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027675,
+      "loss": 2.562,
+      "step": 4190
+    },
+    {
+      "epoch": 6.7056000000000004,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002766,
+      "loss": 3.778,
+      "step": 4191
+    },
+    {
+      "epoch": 6.7072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027644999999999997,
+      "loss": 2.895,
+      "step": 4192
+    },
+    {
+      "epoch": 6.7088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002763,
+      "loss": 4.1391,
+      "step": 4193
+    },
+    {
+      "epoch": 6.7104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027614999999999996,
+      "loss": 3.3999,
+      "step": 4194
+    },
+    {
+      "epoch": 6.712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000276,
+      "loss": 3.3463,
+      "step": 4195
+    },
+    {
+      "epoch": 6.7136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027584999999999996,
+      "loss": 3.2327,
+      "step": 4196
+    },
+    {
+      "epoch": 6.7152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002757,
+      "loss": 2.7379,
+      "step": 4197
+    },
+    {
+      "epoch": 6.7168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027554999999999995,
+      "loss": 3.2524,
+      "step": 4198
+    },
+    {
+      "epoch": 6.7184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027539999999999997,
+      "loss": 3.8617,
+      "step": 4199
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027525,
+      "loss": 3.601,
+      "step": 4200
+    },
+    {
+      "epoch": 6.7216000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027509999999999996,
+      "loss": 4.4073,
+      "step": 4201
+    },
+    {
+      "epoch": 6.7232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027495,
+      "loss": 3.4466,
+      "step": 4202
+    },
+    {
+      "epoch": 6.7248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002748,
+      "loss": 3.2598,
+      "step": 4203
+    },
+    {
+      "epoch": 6.7264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027465,
+      "loss": 3.9411,
+      "step": 4204
+    },
+    {
+      "epoch": 6.728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002745,
+      "loss": 3.5236,
+      "step": 4205
+    },
+    {
+      "epoch": 6.7296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027435,
+      "loss": 3.1362,
+      "step": 4206
+    },
+    {
+      "epoch": 6.7312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002742,
+      "loss": 3.0027,
+      "step": 4207
+    },
+    {
+      "epoch": 6.7328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027404999999999997,
+      "loss": 2.8268,
+      "step": 4208
+    },
+    {
+      "epoch": 6.7344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002739,
+      "loss": 3.9959,
+      "step": 4209
+    },
+    {
+      "epoch": 6.736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027374999999999996,
+      "loss": 2.7458,
+      "step": 4210
+    },
+    {
+      "epoch": 6.7376000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002736,
+      "loss": 3.1352,
+      "step": 4211
+    },
+    {
+      "epoch": 6.7392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027344999999999995,
+      "loss": 3.0027,
+      "step": 4212
+    },
+    {
+      "epoch": 6.7408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002733,
+      "loss": 2.5663,
+      "step": 4213
+    },
+    {
+      "epoch": 6.7424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027314999999999994,
+      "loss": 2.8878,
+      "step": 4214
+    },
+    {
+      "epoch": 6.744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027299999999999997,
+      "loss": 2.4712,
+      "step": 4215
+    },
+    {
+      "epoch": 6.7456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027285,
+      "loss": 2.9749,
+      "step": 4216
+    },
+    {
+      "epoch": 6.7472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027269999999999996,
+      "loss": 3.0081,
+      "step": 4217
+    },
+    {
+      "epoch": 6.7488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027255,
+      "loss": 2.7849,
+      "step": 4218
+    },
+    {
+      "epoch": 6.7504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002724,
+      "loss": 3.5248,
+      "step": 4219
+    },
+    {
+      "epoch": 6.752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027225,
+      "loss": 2.1496,
+      "step": 4220
+    },
+    {
+      "epoch": 6.7536000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002721,
+      "loss": 2.3653,
+      "step": 4221
+    },
+    {
+      "epoch": 6.7552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027194999999999997,
+      "loss": 2.6872,
+      "step": 4222
+    },
+    {
+      "epoch": 6.7568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002718,
+      "loss": 2.7962,
+      "step": 4223
+    },
+    {
+      "epoch": 6.7584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027164999999999996,
+      "loss": 2.6521,
+      "step": 4224
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002715,
+      "loss": 2.9779,
+      "step": 4225
+    },
+    {
+      "epoch": 6.7616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027134999999999995,
+      "loss": 2.8511,
+      "step": 4226
+    },
+    {
+      "epoch": 6.7632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002712,
+      "loss": 3.1687,
+      "step": 4227
+    },
+    {
+      "epoch": 6.7648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027104999999999995,
+      "loss": 2.8407,
+      "step": 4228
+    },
+    {
+      "epoch": 6.7664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027089999999999997,
+      "loss": 2.4979,
+      "step": 4229
+    },
+    {
+      "epoch": 6.768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027074999999999994,
+      "loss": 3.0246,
+      "step": 4230
+    },
+    {
+      "epoch": 6.7696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027059999999999996,
+      "loss": 2.3302,
+      "step": 4231
+    },
+    {
+      "epoch": 6.7712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027045,
+      "loss": 3.2841,
+      "step": 4232
+    },
+    {
+      "epoch": 6.7728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027029999999999996,
+      "loss": 2.9967,
+      "step": 4233
+    },
+    {
+      "epoch": 6.7744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027015,
+      "loss": 3.3507,
+      "step": 4234
+    },
+    {
+      "epoch": 6.776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00027,
+      "loss": 2.5622,
+      "step": 4235
+    },
+    {
+      "epoch": 6.7776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026984999999999997,
+      "loss": 2.9587,
+      "step": 4236
+    },
+    {
+      "epoch": 6.7792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002697,
+      "loss": 3.1645,
+      "step": 4237
+    },
+    {
+      "epoch": 6.7808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026954999999999997,
+      "loss": 2.9535,
+      "step": 4238
+    },
+    {
+      "epoch": 6.7824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002694,
+      "loss": 2.8682,
+      "step": 4239
+    },
+    {
+      "epoch": 6.784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026924999999999996,
+      "loss": 2.3655,
+      "step": 4240
+    },
+    {
+      "epoch": 6.7856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002691,
+      "loss": 3.3084,
+      "step": 4241
+    },
+    {
+      "epoch": 6.7872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026894999999999995,
+      "loss": 2.5359,
+      "step": 4242
+    },
+    {
+      "epoch": 6.7888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002688,
+      "loss": 2.728,
+      "step": 4243
+    },
+    {
+      "epoch": 6.7904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026864999999999994,
+      "loss": 2.4105,
+      "step": 4244
+    },
+    {
+      "epoch": 6.792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026849999999999997,
+      "loss": 2.7605,
+      "step": 4245
+    },
+    {
+      "epoch": 6.7936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026835,
+      "loss": 3.6181,
+      "step": 4246
+    },
+    {
+      "epoch": 6.7952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026819999999999996,
+      "loss": 3.7172,
+      "step": 4247
+    },
+    {
+      "epoch": 6.7968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026805,
+      "loss": 3.3299,
+      "step": 4248
+    },
+    {
+      "epoch": 6.7984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002679,
+      "loss": 2.6778,
+      "step": 4249
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026775,
+      "loss": 3.0789,
+      "step": 4250
+    },
+    {
+      "epoch": 6.8016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002676,
+      "loss": 5.0018,
+      "step": 4251
+    },
+    {
+      "epoch": 6.8032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026744999999999997,
+      "loss": 4.1692,
+      "step": 4252
+    },
+    {
+      "epoch": 6.8048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002673,
+      "loss": 4.7239,
+      "step": 4253
+    },
+    {
+      "epoch": 6.8064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026714999999999996,
+      "loss": 3.4193,
+      "step": 4254
+    },
+    {
+      "epoch": 6.808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000267,
+      "loss": 3.4795,
+      "step": 4255
+    },
+    {
+      "epoch": 6.8096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026684999999999995,
+      "loss": 3.4933,
+      "step": 4256
+    },
+    {
+      "epoch": 6.8112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002667,
+      "loss": 3.6073,
+      "step": 4257
+    },
+    {
+      "epoch": 6.8128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026654999999999995,
+      "loss": 3.0732,
+      "step": 4258
+    },
+    {
+      "epoch": 6.8144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026639999999999997,
+      "loss": 3.1783,
+      "step": 4259
+    },
+    {
+      "epoch": 6.816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026624999999999994,
+      "loss": 3.2414,
+      "step": 4260
+    },
+    {
+      "epoch": 6.8176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026609999999999996,
+      "loss": 3.0902,
+      "step": 4261
+    },
+    {
+      "epoch": 6.8192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026595,
+      "loss": 2.8653,
+      "step": 4262
+    },
+    {
+      "epoch": 6.8208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026579999999999996,
+      "loss": 2.651,
+      "step": 4263
+    },
+    {
+      "epoch": 6.8224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026565,
+      "loss": 3.1109,
+      "step": 4264
+    },
+    {
+      "epoch": 6.824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002655,
+      "loss": 3.5091,
+      "step": 4265
+    },
+    {
+      "epoch": 6.8256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026534999999999997,
+      "loss": 2.6144,
+      "step": 4266
+    },
+    {
+      "epoch": 6.8272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002652,
+      "loss": 2.7945,
+      "step": 4267
+    },
+    {
+      "epoch": 6.8288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026504999999999996,
+      "loss": 2.8182,
+      "step": 4268
+    },
+    {
+      "epoch": 6.8304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002649,
+      "loss": 2.7363,
+      "step": 4269
+    },
+    {
+      "epoch": 6.832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026474999999999996,
+      "loss": 3.0435,
+      "step": 4270
+    },
+    {
+      "epoch": 6.8336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002646,
+      "loss": 2.7486,
+      "step": 4271
+    },
+    {
+      "epoch": 6.8352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026444999999999995,
+      "loss": 2.567,
+      "step": 4272
+    },
+    {
+      "epoch": 6.8368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002643,
+      "loss": 3.1096,
+      "step": 4273
+    },
+    {
+      "epoch": 6.8384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026414999999999994,
+      "loss": 2.5198,
+      "step": 4274
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 3.0245,
+      "step": 4275
+    },
+    {
+      "epoch": 6.8416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026384999999999994,
+      "loss": 3.0371,
+      "step": 4276
+    },
+    {
+      "epoch": 6.8431999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026369999999999996,
+      "loss": 2.8618,
+      "step": 4277
+    },
+    {
+      "epoch": 6.8448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026355,
+      "loss": 2.5902,
+      "step": 4278
+    },
+    {
+      "epoch": 6.8464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026339999999999995,
+      "loss": 2.687,
+      "step": 4279
+    },
+    {
+      "epoch": 6.848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026325,
+      "loss": 2.7124,
+      "step": 4280
+    },
+    {
+      "epoch": 6.8496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002631,
+      "loss": 3.6682,
+      "step": 4281
+    },
+    {
+      "epoch": 6.8512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026294999999999997,
+      "loss": 2.4604,
+      "step": 4282
+    },
+    {
+      "epoch": 6.8528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002628,
+      "loss": 2.8002,
+      "step": 4283
+    },
+    {
+      "epoch": 6.8544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026264999999999996,
+      "loss": 2.6395,
+      "step": 4284
+    },
+    {
+      "epoch": 6.856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002625,
+      "loss": 2.4311,
+      "step": 4285
+    },
+    {
+      "epoch": 6.8576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026235,
+      "loss": 2.9462,
+      "step": 4286
+    },
+    {
+      "epoch": 6.8591999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002622,
+      "loss": 3.6116,
+      "step": 4287
+    },
+    {
+      "epoch": 6.8608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026205,
+      "loss": 3.5986,
+      "step": 4288
+    },
+    {
+      "epoch": 6.8624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026189999999999997,
+      "loss": 2.9106,
+      "step": 4289
+    },
+    {
+      "epoch": 6.864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026175,
+      "loss": 3.1669,
+      "step": 4290
+    },
+    {
+      "epoch": 6.8656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026159999999999996,
+      "loss": 3.6225,
+      "step": 4291
+    },
+    {
+      "epoch": 6.8672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026145,
+      "loss": 3.0465,
+      "step": 4292
+    },
+    {
+      "epoch": 6.8688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026129999999999995,
+      "loss": 3.0488,
+      "step": 4293
+    },
+    {
+      "epoch": 6.8704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026115,
+      "loss": 2.2928,
+      "step": 4294
+    },
+    {
+      "epoch": 6.872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000261,
+      "loss": 2.9981,
+      "step": 4295
+    },
+    {
+      "epoch": 6.8736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026084999999999997,
+      "loss": 2.8215,
+      "step": 4296
+    },
+    {
+      "epoch": 6.8751999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002607,
+      "loss": 2.9482,
+      "step": 4297
+    },
+    {
+      "epoch": 6.8768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026055,
+      "loss": 2.9353,
+      "step": 4298
+    },
+    {
+      "epoch": 6.8784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002604,
+      "loss": 2.7093,
+      "step": 4299
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00026025,
+      "loss": 5.2044,
+      "step": 4300
+    },
+    {
+      "epoch": 6.8816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002601,
+      "loss": 4.0005,
+      "step": 4301
+    },
+    {
+      "epoch": 6.8832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025995,
+      "loss": 3.5939,
+      "step": 4302
+    },
+    {
+      "epoch": 6.8848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025979999999999997,
+      "loss": 4.2119,
+      "step": 4303
+    },
+    {
+      "epoch": 6.8864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025965,
+      "loss": 3.7944,
+      "step": 4304
+    },
+    {
+      "epoch": 6.888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025949999999999997,
+      "loss": 3.7665,
+      "step": 4305
+    },
+    {
+      "epoch": 6.8896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025935,
+      "loss": 3.5207,
+      "step": 4306
+    },
+    {
+      "epoch": 6.8911999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025919999999999996,
+      "loss": 2.9676,
+      "step": 4307
+    },
+    {
+      "epoch": 6.8928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025905,
+      "loss": 2.6527,
+      "step": 4308
+    },
+    {
+      "epoch": 6.8944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025889999999999995,
+      "loss": 3.0158,
+      "step": 4309
+    },
+    {
+      "epoch": 6.896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025875,
+      "loss": 2.931,
+      "step": 4310
+    },
+    {
+      "epoch": 6.8976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002586,
+      "loss": 3.003,
+      "step": 4311
+    },
+    {
+      "epoch": 6.8992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025844999999999997,
+      "loss": 3.6457,
+      "step": 4312
+    },
+    {
+      "epoch": 6.9008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002583,
+      "loss": 2.887,
+      "step": 4313
+    },
+    {
+      "epoch": 6.9024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025815,
+      "loss": 2.5508,
+      "step": 4314
+    },
+    {
+      "epoch": 6.904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000258,
+      "loss": 2.8492,
+      "step": 4315
+    },
+    {
+      "epoch": 6.9056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025785,
+      "loss": 2.5418,
+      "step": 4316
+    },
+    {
+      "epoch": 6.9072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002577,
+      "loss": 2.7807,
+      "step": 4317
+    },
+    {
+      "epoch": 6.9088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025755,
+      "loss": 2.6149,
+      "step": 4318
+    },
+    {
+      "epoch": 6.9104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025739999999999997,
+      "loss": 2.6781,
+      "step": 4319
+    },
+    {
+      "epoch": 6.912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025725,
+      "loss": 2.3818,
+      "step": 4320
+    },
+    {
+      "epoch": 6.9136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025709999999999996,
+      "loss": 3.4684,
+      "step": 4321
+    },
+    {
+      "epoch": 6.9152000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025695,
+      "loss": 2.6856,
+      "step": 4322
+    },
+    {
+      "epoch": 6.9168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025679999999999995,
+      "loss": 3.0432,
+      "step": 4323
+    },
+    {
+      "epoch": 6.9184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025665,
+      "loss": 2.5531,
+      "step": 4324
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025649999999999995,
+      "loss": 2.0887,
+      "step": 4325
+    },
+    {
+      "epoch": 6.9216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025634999999999997,
+      "loss": 3.3472,
+      "step": 4326
+    },
+    {
+      "epoch": 6.9232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002562,
+      "loss": 2.7786,
+      "step": 4327
+    },
+    {
+      "epoch": 6.9248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025604999999999996,
+      "loss": 3.0112,
+      "step": 4328
+    },
+    {
+      "epoch": 6.9264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002559,
+      "loss": 3.2129,
+      "step": 4329
+    },
+    {
+      "epoch": 6.928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025575,
+      "loss": 3.3005,
+      "step": 4330
+    },
+    {
+      "epoch": 6.9296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002556,
+      "loss": 2.9553,
+      "step": 4331
+    },
+    {
+      "epoch": 6.9312000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025545,
+      "loss": 3.0149,
+      "step": 4332
+    },
+    {
+      "epoch": 6.9328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025529999999999997,
+      "loss": 2.7872,
+      "step": 4333
+    },
+    {
+      "epoch": 6.9344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025515,
+      "loss": 2.9444,
+      "step": 4334
+    },
+    {
+      "epoch": 6.936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 2.7883,
+      "step": 4335
+    },
+    {
+      "epoch": 6.9376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025485,
+      "loss": 2.5147,
+      "step": 4336
+    },
+    {
+      "epoch": 6.9392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025469999999999996,
+      "loss": 2.6618,
+      "step": 4337
+    },
+    {
+      "epoch": 6.9408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025455,
+      "loss": 3.1769,
+      "step": 4338
+    },
+    {
+      "epoch": 6.9424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025439999999999995,
+      "loss": 2.593,
+      "step": 4339
+    },
+    {
+      "epoch": 6.944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025425,
+      "loss": 2.4697,
+      "step": 4340
+    },
+    {
+      "epoch": 6.9456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002541,
+      "loss": 2.3677,
+      "step": 4341
+    },
+    {
+      "epoch": 6.9472000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025394999999999997,
+      "loss": 3.4864,
+      "step": 4342
+    },
+    {
+      "epoch": 6.9488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002538,
+      "loss": 2.7577,
+      "step": 4343
+    },
+    {
+      "epoch": 6.9504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025365,
+      "loss": 2.7377,
+      "step": 4344
+    },
+    {
+      "epoch": 6.952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002535,
+      "loss": 2.7041,
+      "step": 4345
+    },
+    {
+      "epoch": 6.9536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025335,
+      "loss": 2.4056,
+      "step": 4346
+    },
+    {
+      "epoch": 6.9552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002532,
+      "loss": 3.782,
+      "step": 4347
+    },
+    {
+      "epoch": 6.9568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025305,
+      "loss": 3.3568,
+      "step": 4348
+    },
+    {
+      "epoch": 6.9584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025289999999999997,
+      "loss": 4.0026,
+      "step": 4349
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": NaN,
+      "learning_rate": 0.00025289999999999997,
+      "loss": 3.1076,
+      "step": 4350
+    },
+    {
+      "epoch": 6.9616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025275,
+      "loss": 4.3278,
+      "step": 4351
+    },
+    {
+      "epoch": 6.9632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025259999999999996,
+      "loss": 3.6904,
+      "step": 4352
+    },
+    {
+      "epoch": 6.9648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025245,
+      "loss": 2.9045,
+      "step": 4353
+    },
+    {
+      "epoch": 6.9664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025229999999999995,
+      "loss": 4.1884,
+      "step": 4354
+    },
+    {
+      "epoch": 6.968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025215,
+      "loss": 4.3218,
+      "step": 4355
+    },
+    {
+      "epoch": 6.9696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 2.6043,
+      "step": 4356
+    },
+    {
+      "epoch": 6.9712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025184999999999997,
+      "loss": 2.6978,
+      "step": 4357
+    },
+    {
+      "epoch": 6.9728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002517,
+      "loss": 2.8147,
+      "step": 4358
+    },
+    {
+      "epoch": 6.9744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025154999999999996,
+      "loss": 2.6158,
+      "step": 4359
+    },
+    {
+      "epoch": 6.976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002514,
+      "loss": 2.6996,
+      "step": 4360
+    },
+    {
+      "epoch": 6.9776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025125,
+      "loss": 3.5713,
+      "step": 4361
+    },
+    {
+      "epoch": 6.9792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002511,
+      "loss": 3.194,
+      "step": 4362
+    },
+    {
+      "epoch": 6.9808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025095,
+      "loss": 3.3608,
+      "step": 4363
+    },
+    {
+      "epoch": 6.9824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025079999999999997,
+      "loss": 2.3152,
+      "step": 4364
+    },
+    {
+      "epoch": 6.984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025065,
+      "loss": 2.8111,
+      "step": 4365
+    },
+    {
+      "epoch": 6.9856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025049999999999996,
+      "loss": 3.17,
+      "step": 4366
+    },
+    {
+      "epoch": 6.9872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025035,
+      "loss": 2.7564,
+      "step": 4367
+    },
+    {
+      "epoch": 6.9888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025019999999999996,
+      "loss": 3.6902,
+      "step": 4368
+    },
+    {
+      "epoch": 6.9904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00025005,
+      "loss": 2.6898,
+      "step": 4369
+    },
+    {
+      "epoch": 6.992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024989999999999995,
+      "loss": 3.0014,
+      "step": 4370
+    },
+    {
+      "epoch": 6.9936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024974999999999997,
+      "loss": 2.8975,
+      "step": 4371
+    },
+    {
+      "epoch": 6.9952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024959999999999994,
+      "loss": 2.2888,
+      "step": 4372
+    },
+    {
+      "epoch": 6.9968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024944999999999996,
+      "loss": 3.0213,
+      "step": 4373
+    },
+    {
+      "epoch": 6.9984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002493,
+      "loss": 2.0513,
+      "step": 4374
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024914999999999996,
+      "loss": 3.1802,
+      "step": 4375
+    },
+    {
+      "epoch": 7.0016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000249,
+      "loss": 4.4198,
+      "step": 4376
+    },
+    {
+      "epoch": 7.0032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024885,
+      "loss": 4.4398,
+      "step": 4377
+    },
+    {
+      "epoch": 7.0048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002487,
+      "loss": 3.1124,
+      "step": 4378
+    },
+    {
+      "epoch": 7.0064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024855,
+      "loss": 4.2911,
+      "step": 4379
+    },
+    {
+      "epoch": 7.008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024839999999999997,
+      "loss": 4.1504,
+      "step": 4380
+    },
+    {
+      "epoch": 7.0096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024825,
+      "loss": 2.849,
+      "step": 4381
+    },
+    {
+      "epoch": 7.0112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024809999999999996,
+      "loss": 2.7981,
+      "step": 4382
+    },
+    {
+      "epoch": 7.0128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024795,
+      "loss": 3.4686,
+      "step": 4383
+    },
+    {
+      "epoch": 7.0144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024779999999999995,
+      "loss": 3.4509,
+      "step": 4384
+    },
+    {
+      "epoch": 7.016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024765,
+      "loss": 3.4504,
+      "step": 4385
+    },
+    {
+      "epoch": 7.0176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024749999999999994,
+      "loss": 2.8491,
+      "step": 4386
+    },
+    {
+      "epoch": 7.0192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024734999999999997,
+      "loss": 3.495,
+      "step": 4387
+    },
+    {
+      "epoch": 7.0208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002472,
+      "loss": 2.9273,
+      "step": 4388
+    },
+    {
+      "epoch": 7.0224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024704999999999996,
+      "loss": 3.0325,
+      "step": 4389
+    },
+    {
+      "epoch": 7.024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002469,
+      "loss": 3.1798,
+      "step": 4390
+    },
+    {
+      "epoch": 7.0256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024675,
+      "loss": 3.3611,
+      "step": 4391
+    },
+    {
+      "epoch": 7.0272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002466,
+      "loss": 2.7285,
+      "step": 4392
+    },
+    {
+      "epoch": 7.0288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024645,
+      "loss": 2.5132,
+      "step": 4393
+    },
+    {
+      "epoch": 7.0304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024629999999999997,
+      "loss": 2.9996,
+      "step": 4394
+    },
+    {
+      "epoch": 7.032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024615,
+      "loss": 2.6329,
+      "step": 4395
+    },
+    {
+      "epoch": 7.0336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 2.4842,
+      "step": 4396
+    },
+    {
+      "epoch": 7.0352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024585,
+      "loss": 2.9166,
+      "step": 4397
+    },
+    {
+      "epoch": 7.0368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024569999999999995,
+      "loss": 2.2728,
+      "step": 4398
+    },
+    {
+      "epoch": 7.0384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024555,
+      "loss": 3.182,
+      "step": 4399
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024539999999999995,
+      "loss": 3.4259,
+      "step": 4400
+    },
+    {
+      "epoch": 7.0416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024524999999999997,
+      "loss": 2.6928,
+      "step": 4401
+    },
+    {
+      "epoch": 7.0432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024509999999999994,
+      "loss": 2.9651,
+      "step": 4402
+    },
+    {
+      "epoch": 7.0448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024494999999999996,
+      "loss": 2.5092,
+      "step": 4403
+    },
+    {
+      "epoch": 7.0464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002448,
+      "loss": 2.5621,
+      "step": 4404
+    },
+    {
+      "epoch": 7.048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024464999999999996,
+      "loss": 2.1604,
+      "step": 4405
+    },
+    {
+      "epoch": 7.0496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002445,
+      "loss": 2.8104,
+      "step": 4406
+    },
+    {
+      "epoch": 7.0512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024435,
+      "loss": 3.2445,
+      "step": 4407
+    },
+    {
+      "epoch": 7.0528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024419999999999997,
+      "loss": 3.0785,
+      "step": 4408
+    },
+    {
+      "epoch": 7.0544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024404999999999997,
+      "loss": 3.4991,
+      "step": 4409
+    },
+    {
+      "epoch": 7.056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024389999999999997,
+      "loss": 2.883,
+      "step": 4410
+    },
+    {
+      "epoch": 7.0576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024375,
+      "loss": 2.8117,
+      "step": 4411
+    },
+    {
+      "epoch": 7.0592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024359999999999999,
+      "loss": 3.6825,
+      "step": 4412
+    },
+    {
+      "epoch": 7.0608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024344999999999998,
+      "loss": 2.6544,
+      "step": 4413
+    },
+    {
+      "epoch": 7.0624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002433,
+      "loss": 2.7173,
+      "step": 4414
+    },
+    {
+      "epoch": 7.064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024314999999999997,
+      "loss": 3.1109,
+      "step": 4415
+    },
+    {
+      "epoch": 7.0656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000243,
+      "loss": 3.0858,
+      "step": 4416
+    },
+    {
+      "epoch": 7.0672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024284999999999997,
+      "loss": 2.5237,
+      "step": 4417
+    },
+    {
+      "epoch": 7.0688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002427,
+      "loss": 2.5561,
+      "step": 4418
+    },
+    {
+      "epoch": 7.0704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024255,
+      "loss": 3.4563,
+      "step": 4419
+    },
+    {
+      "epoch": 7.072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024239999999999998,
+      "loss": 2.5289,
+      "step": 4420
+    },
+    {
+      "epoch": 7.0736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024224999999999998,
+      "loss": 3.3271,
+      "step": 4421
+    },
+    {
+      "epoch": 7.0752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002421,
+      "loss": 3.2885,
+      "step": 4422
+    },
+    {
+      "epoch": 7.0768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024194999999999997,
+      "loss": 3.2276,
+      "step": 4423
+    },
+    {
+      "epoch": 7.0784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002418,
+      "loss": 2.785,
+      "step": 4424
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024164999999999996,
+      "loss": 3.4965,
+      "step": 4425
+    },
+    {
+      "epoch": 7.0816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002415,
+      "loss": 4.8508,
+      "step": 4426
+    },
+    {
+      "epoch": 7.0832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024134999999999998,
+      "loss": 2.9179,
+      "step": 4427
+    },
+    {
+      "epoch": 7.0848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024119999999999998,
+      "loss": 4.5402,
+      "step": 4428
+    },
+    {
+      "epoch": 7.0864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024104999999999998,
+      "loss": 3.5119,
+      "step": 4429
+    },
+    {
+      "epoch": 7.088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002409,
+      "loss": 3.4308,
+      "step": 4430
+    },
+    {
+      "epoch": 7.0896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024074999999999997,
+      "loss": 2.709,
+      "step": 4431
+    },
+    {
+      "epoch": 7.0912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002406,
+      "loss": 3.1521,
+      "step": 4432
+    },
+    {
+      "epoch": 7.0928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024044999999999996,
+      "loss": 2.9411,
+      "step": 4433
+    },
+    {
+      "epoch": 7.0944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024029999999999999,
+      "loss": 3.0063,
+      "step": 4434
+    },
+    {
+      "epoch": 7.096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00024014999999999998,
+      "loss": 3.4764,
+      "step": 4435
+    },
+    {
+      "epoch": 7.0976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 3.1709,
+      "step": 4436
+    },
+    {
+      "epoch": 7.0992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023984999999999998,
+      "loss": 2.8713,
+      "step": 4437
+    },
+    {
+      "epoch": 7.1008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002397,
+      "loss": 2.8115,
+      "step": 4438
+    },
+    {
+      "epoch": 7.1024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023954999999999997,
+      "loss": 2.7006,
+      "step": 4439
+    },
+    {
+      "epoch": 7.104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002394,
+      "loss": 3.5017,
+      "step": 4440
+    },
+    {
+      "epoch": 7.1056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023925,
+      "loss": 3.0051,
+      "step": 4441
+    },
+    {
+      "epoch": 7.1072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023909999999999998,
+      "loss": 3.0562,
+      "step": 4442
+    },
+    {
+      "epoch": 7.1088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023894999999999998,
+      "loss": 2.5437,
+      "step": 4443
+    },
+    {
+      "epoch": 7.1104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002388,
+      "loss": 2.7862,
+      "step": 4444
+    },
+    {
+      "epoch": 7.112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023864999999999997,
+      "loss": 2.838,
+      "step": 4445
+    },
+    {
+      "epoch": 7.1136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002385,
+      "loss": 2.2052,
+      "step": 4446
+    },
+    {
+      "epoch": 7.1152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023834999999999997,
+      "loss": 2.9081,
+      "step": 4447
+    },
+    {
+      "epoch": 7.1168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002382,
+      "loss": 2.4304,
+      "step": 4448
+    },
+    {
+      "epoch": 7.1184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023804999999999999,
+      "loss": 2.941,
+      "step": 4449
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023789999999999998,
+      "loss": 2.7017,
+      "step": 4450
+    },
+    {
+      "epoch": 7.1216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023774999999999998,
+      "loss": 2.8525,
+      "step": 4451
+    },
+    {
+      "epoch": 7.1232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002376,
+      "loss": 3.0987,
+      "step": 4452
+    },
+    {
+      "epoch": 7.1248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023744999999999997,
+      "loss": 2.4817,
+      "step": 4453
+    },
+    {
+      "epoch": 7.1264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002373,
+      "loss": 3.2059,
+      "step": 4454
+    },
+    {
+      "epoch": 7.128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023714999999999996,
+      "loss": 2.5229,
+      "step": 4455
+    },
+    {
+      "epoch": 7.1296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000237,
+      "loss": 2.9222,
+      "step": 4456
+    },
+    {
+      "epoch": 7.1312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023684999999999998,
+      "loss": 3.0885,
+      "step": 4457
+    },
+    {
+      "epoch": 7.1328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023669999999999998,
+      "loss": 3.4679,
+      "step": 4458
+    },
+    {
+      "epoch": 7.1344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023654999999999998,
+      "loss": 2.9408,
+      "step": 4459
+    },
+    {
+      "epoch": 7.136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002364,
+      "loss": 2.8065,
+      "step": 4460
+    },
+    {
+      "epoch": 7.1376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023624999999999997,
+      "loss": 3.0352,
+      "step": 4461
+    },
+    {
+      "epoch": 7.1392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002361,
+      "loss": 2.9077,
+      "step": 4462
+    },
+    {
+      "epoch": 7.1408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023594999999999996,
+      "loss": 3.1758,
+      "step": 4463
+    },
+    {
+      "epoch": 7.1424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023579999999999999,
+      "loss": 4.3853,
+      "step": 4464
+    },
+    {
+      "epoch": 7.144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023564999999999998,
+      "loss": 3.2723,
+      "step": 4465
+    },
+    {
+      "epoch": 7.1456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023549999999999998,
+      "loss": 4.1234,
+      "step": 4466
+    },
+    {
+      "epoch": 7.1472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023534999999999997,
+      "loss": 2.7587,
+      "step": 4467
+    },
+    {
+      "epoch": 7.1488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002352,
+      "loss": 3.2122,
+      "step": 4468
+    },
+    {
+      "epoch": 7.1504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023504999999999997,
+      "loss": 3.7312,
+      "step": 4469
+    },
+    {
+      "epoch": 7.152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002349,
+      "loss": 3.0445,
+      "step": 4470
+    },
+    {
+      "epoch": 7.1536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023474999999999996,
+      "loss": 2.5092,
+      "step": 4471
+    },
+    {
+      "epoch": 7.1552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023459999999999998,
+      "loss": 2.8707,
+      "step": 4472
+    },
+    {
+      "epoch": 7.1568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023444999999999998,
+      "loss": 3.4896,
+      "step": 4473
+    },
+    {
+      "epoch": 7.1584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023429999999999998,
+      "loss": 2.7866,
+      "step": 4474
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": NaN,
+      "learning_rate": 0.00023429999999999998,
+      "loss": 3.4577,
+      "step": 4475
+    },
+    {
+      "epoch": 7.1616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023414999999999997,
+      "loss": 4.0787,
+      "step": 4476
+    },
+    {
+      "epoch": 7.1632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000234,
+      "loss": 3.5122,
+      "step": 4477
+    },
+    {
+      "epoch": 7.1648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023384999999999997,
+      "loss": 4.6246,
+      "step": 4478
+    },
+    {
+      "epoch": 7.1664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002337,
+      "loss": 3.394,
+      "step": 4479
+    },
+    {
+      "epoch": 7.168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023354999999999996,
+      "loss": 3.8507,
+      "step": 4480
+    },
+    {
+      "epoch": 7.1696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023339999999999998,
+      "loss": 4.4262,
+      "step": 4481
+    },
+    {
+      "epoch": 7.1712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023324999999999998,
+      "loss": 3.6319,
+      "step": 4482
+    },
+    {
+      "epoch": 7.1728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023309999999999997,
+      "loss": 3.8157,
+      "step": 4483
+    },
+    {
+      "epoch": 7.1744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023294999999999997,
+      "loss": 3.4546,
+      "step": 4484
+    },
+    {
+      "epoch": 7.176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002328,
+      "loss": 2.7881,
+      "step": 4485
+    },
+    {
+      "epoch": 7.1776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023264999999999996,
+      "loss": 3.051,
+      "step": 4486
+    },
+    {
+      "epoch": 7.1792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023249999999999999,
+      "loss": 3.8072,
+      "step": 4487
+    },
+    {
+      "epoch": 7.1808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023234999999999998,
+      "loss": 3.5231,
+      "step": 4488
+    },
+    {
+      "epoch": 7.1824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023219999999999998,
+      "loss": 2.9368,
+      "step": 4489
+    },
+    {
+      "epoch": 7.184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023204999999999998,
+      "loss": 2.2729,
+      "step": 4490
+    },
+    {
+      "epoch": 7.1856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002319,
+      "loss": 2.8623,
+      "step": 4491
+    },
+    {
+      "epoch": 7.1872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023174999999999997,
+      "loss": 2.7405,
+      "step": 4492
+    },
+    {
+      "epoch": 7.1888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002316,
+      "loss": 3.0231,
+      "step": 4493
+    },
+    {
+      "epoch": 7.1904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023144999999999996,
+      "loss": 2.6991,
+      "step": 4494
+    },
+    {
+      "epoch": 7.192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023129999999999998,
+      "loss": 2.6831,
+      "step": 4495
+    },
+    {
+      "epoch": 7.1936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023114999999999998,
+      "loss": 2.8213,
+      "step": 4496
+    },
+    {
+      "epoch": 7.1952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023099999999999998,
+      "loss": 2.3582,
+      "step": 4497
+    },
+    {
+      "epoch": 7.1968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023084999999999997,
+      "loss": 4.7609,
+      "step": 4498
+    },
+    {
+      "epoch": 7.1984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002307,
+      "loss": 2.6573,
+      "step": 4499
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023054999999999997,
+      "loss": 2.6968,
+      "step": 4500
+    },
+    {
+      "epoch": 7.2016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002304,
+      "loss": 2.8074,
+      "step": 4501
+    },
+    {
+      "epoch": 7.2032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023024999999999996,
+      "loss": 3.4635,
+      "step": 4502
+    },
+    {
+      "epoch": 7.2048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00023009999999999998,
+      "loss": 3.3308,
+      "step": 4503
+    },
+    {
+      "epoch": 7.2064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022994999999999998,
+      "loss": 2.6982,
+      "step": 4504
+    },
+    {
+      "epoch": 7.208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022979999999999997,
+      "loss": 2.4931,
+      "step": 4505
+    },
+    {
+      "epoch": 7.2096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022964999999999997,
+      "loss": 3.8077,
+      "step": 4506
+    },
+    {
+      "epoch": 7.2112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002295,
+      "loss": 2.4663,
+      "step": 4507
+    },
+    {
+      "epoch": 7.2128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022934999999999996,
+      "loss": 2.4444,
+      "step": 4508
+    },
+    {
+      "epoch": 7.2144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002292,
+      "loss": 2.8851,
+      "step": 4509
+    },
+    {
+      "epoch": 7.216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022904999999999996,
+      "loss": 3.0816,
+      "step": 4510
+    },
+    {
+      "epoch": 7.2176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022889999999999998,
+      "loss": 2.8437,
+      "step": 4511
+    },
+    {
+      "epoch": 7.2192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022874999999999998,
+      "loss": 2.5104,
+      "step": 4512
+    },
+    {
+      "epoch": 7.2208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022859999999999997,
+      "loss": 3.1474,
+      "step": 4513
+    },
+    {
+      "epoch": 7.2224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022844999999999997,
+      "loss": 3.1323,
+      "step": 4514
+    },
+    {
+      "epoch": 7.224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002283,
+      "loss": 3.3355,
+      "step": 4515
+    },
+    {
+      "epoch": 7.2256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022814999999999996,
+      "loss": 2.3959,
+      "step": 4516
+    },
+    {
+      "epoch": 7.2272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 3.0254,
+      "step": 4517
+    },
+    {
+      "epoch": 7.2288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022784999999999995,
+      "loss": 2.8069,
+      "step": 4518
+    },
+    {
+      "epoch": 7.2304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022769999999999998,
+      "loss": 3.5841,
+      "step": 4519
+    },
+    {
+      "epoch": 7.232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022754999999999997,
+      "loss": 2.5879,
+      "step": 4520
+    },
+    {
+      "epoch": 7.2336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022739999999999997,
+      "loss": 3.2762,
+      "step": 4521
+    },
+    {
+      "epoch": 7.2352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022724999999999997,
+      "loss": 2.0486,
+      "step": 4522
+    },
+    {
+      "epoch": 7.2368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002271,
+      "loss": 2.6137,
+      "step": 4523
+    },
+    {
+      "epoch": 7.2384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022694999999999996,
+      "loss": 3.9723,
+      "step": 4524
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022679999999999998,
+      "loss": 4.651,
+      "step": 4525
+    },
+    {
+      "epoch": 7.2416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022664999999999995,
+      "loss": 4.3543,
+      "step": 4526
+    },
+    {
+      "epoch": 7.2432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022649999999999998,
+      "loss": 3.9596,
+      "step": 4527
+    },
+    {
+      "epoch": 7.2448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022634999999999997,
+      "loss": 3.3218,
+      "step": 4528
+    },
+    {
+      "epoch": 7.2464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022619999999999997,
+      "loss": 3.7675,
+      "step": 4529
+    },
+    {
+      "epoch": 7.248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022604999999999997,
+      "loss": 3.9477,
+      "step": 4530
+    },
+    {
+      "epoch": 7.2496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002259,
+      "loss": 2.8677,
+      "step": 4531
+    },
+    {
+      "epoch": 7.2512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022574999999999996,
+      "loss": 3.0615,
+      "step": 4532
+    },
+    {
+      "epoch": 7.2528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022559999999999998,
+      "loss": 3.2326,
+      "step": 4533
+    },
+    {
+      "epoch": 7.2544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022544999999999995,
+      "loss": 2.5745,
+      "step": 4534
+    },
+    {
+      "epoch": 7.256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022529999999999997,
+      "loss": 3.1765,
+      "step": 4535
+    },
+    {
+      "epoch": 7.2576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022514999999999997,
+      "loss": 3.271,
+      "step": 4536
+    },
+    {
+      "epoch": 7.2592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000225,
+      "loss": 3.1558,
+      "step": 4537
+    },
+    {
+      "epoch": 7.2608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022485,
+      "loss": 2.5605,
+      "step": 4538
+    },
+    {
+      "epoch": 7.2624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002247,
+      "loss": 2.6944,
+      "step": 4539
+    },
+    {
+      "epoch": 7.264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022455,
+      "loss": 2.6869,
+      "step": 4540
+    },
+    {
+      "epoch": 7.2656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022439999999999998,
+      "loss": 2.8473,
+      "step": 4541
+    },
+    {
+      "epoch": 7.2672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022425,
+      "loss": 2.8917,
+      "step": 4542
+    },
+    {
+      "epoch": 7.2688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022409999999999997,
+      "loss": 2.9746,
+      "step": 4543
+    },
+    {
+      "epoch": 7.2704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022395,
+      "loss": 3.6293,
+      "step": 4544
+    },
+    {
+      "epoch": 7.272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002238,
+      "loss": 2.8313,
+      "step": 4545
+    },
+    {
+      "epoch": 7.2736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022365,
+      "loss": 2.3467,
+      "step": 4546
+    },
+    {
+      "epoch": 7.2752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022349999999999998,
+      "loss": 2.8612,
+      "step": 4547
+    },
+    {
+      "epoch": 7.2768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022335,
+      "loss": 2.7406,
+      "step": 4548
+    },
+    {
+      "epoch": 7.2783999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022319999999999998,
+      "loss": 2.6151,
+      "step": 4549
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022305,
+      "loss": 2.3726,
+      "step": 4550
+    },
+    {
+      "epoch": 7.2816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022289999999999997,
+      "loss": 2.4332,
+      "step": 4551
+    },
+    {
+      "epoch": 7.2832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022275,
+      "loss": 2.9835,
+      "step": 4552
+    },
+    {
+      "epoch": 7.2848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002226,
+      "loss": 2.5813,
+      "step": 4553
+    },
+    {
+      "epoch": 7.2864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022244999999999999,
+      "loss": 2.9601,
+      "step": 4554
+    },
+    {
+      "epoch": 7.288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022229999999999998,
+      "loss": 2.7316,
+      "step": 4555
+    },
+    {
+      "epoch": 7.2896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022215,
+      "loss": 2.8432,
+      "step": 4556
+    },
+    {
+      "epoch": 7.2912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 2.6512,
+      "step": 4557
+    },
+    {
+      "epoch": 7.2928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022185,
+      "loss": 2.9372,
+      "step": 4558
+    },
+    {
+      "epoch": 7.2943999999999996,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022169999999999997,
+      "loss": 2.3597,
+      "step": 4559
+    },
+    {
+      "epoch": 7.296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022155,
+      "loss": 2.6766,
+      "step": 4560
+    },
+    {
+      "epoch": 7.2976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002214,
+      "loss": 2.6093,
+      "step": 4561
+    },
+    {
+      "epoch": 7.2992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022124999999999998,
+      "loss": 3.3046,
+      "step": 4562
+    },
+    {
+      "epoch": 7.3008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022109999999999998,
+      "loss": 2.4641,
+      "step": 4563
+    },
+    {
+      "epoch": 7.3024000000000004,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022095,
+      "loss": 3.1234,
+      "step": 4564
+    },
+    {
+      "epoch": 7.304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022079999999999997,
+      "loss": 2.3525,
+      "step": 4565
+    },
+    {
+      "epoch": 7.3056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022065,
+      "loss": 3.0467,
+      "step": 4566
+    },
+    {
+      "epoch": 7.3072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022049999999999997,
+      "loss": 2.6771,
+      "step": 4567
+    },
+    {
+      "epoch": 7.3088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022035,
+      "loss": 3.5626,
+      "step": 4568
+    },
+    {
+      "epoch": 7.3104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022019999999999999,
+      "loss": 3.0666,
+      "step": 4569
+    },
+    {
+      "epoch": 7.312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00022004999999999998,
+      "loss": 3.374,
+      "step": 4570
+    },
+    {
+      "epoch": 7.3136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021989999999999998,
+      "loss": 3.1645,
+      "step": 4571
+    },
+    {
+      "epoch": 7.3152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021975,
+      "loss": 2.5583,
+      "step": 4572
+    },
+    {
+      "epoch": 7.3168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021959999999999997,
+      "loss": 3.6608,
+      "step": 4573
+    },
+    {
+      "epoch": 7.3184000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021945,
+      "loss": 4.1134,
+      "step": 4574
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021929999999999996,
+      "loss": 3.1932,
+      "step": 4575
+    },
+    {
+      "epoch": 7.3216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021915,
+      "loss": 3.548,
+      "step": 4576
+    },
+    {
+      "epoch": 7.3232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 3.1762,
+      "step": 4577
+    },
+    {
+      "epoch": 7.3248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021884999999999998,
+      "loss": 4.0924,
+      "step": 4578
+    },
+    {
+      "epoch": 7.3264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021869999999999998,
+      "loss": 3.4292,
+      "step": 4579
+    },
+    {
+      "epoch": 7.328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021855,
+      "loss": 3.931,
+      "step": 4580
+    },
+    {
+      "epoch": 7.3296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021839999999999997,
+      "loss": 2.9257,
+      "step": 4581
+    },
+    {
+      "epoch": 7.3312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021825,
+      "loss": 3.6792,
+      "step": 4582
+    },
+    {
+      "epoch": 7.3328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021809999999999996,
+      "loss": 3.305,
+      "step": 4583
+    },
+    {
+      "epoch": 7.3344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021794999999999999,
+      "loss": 2.4825,
+      "step": 4584
+    },
+    {
+      "epoch": 7.336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021779999999999998,
+      "loss": 3.573,
+      "step": 4585
+    },
+    {
+      "epoch": 7.3376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021764999999999998,
+      "loss": 2.5883,
+      "step": 4586
+    },
+    {
+      "epoch": 7.3392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021749999999999997,
+      "loss": 3.0585,
+      "step": 4587
+    },
+    {
+      "epoch": 7.3408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021735,
+      "loss": 2.9204,
+      "step": 4588
+    },
+    {
+      "epoch": 7.3424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021719999999999997,
+      "loss": 2.9738,
+      "step": 4589
+    },
+    {
+      "epoch": 7.344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021705,
+      "loss": 2.8103,
+      "step": 4590
+    },
+    {
+      "epoch": 7.3456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002169,
+      "loss": 2.9121,
+      "step": 4591
+    },
+    {
+      "epoch": 7.3472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021674999999999998,
+      "loss": 2.4368,
+      "step": 4592
+    },
+    {
+      "epoch": 7.3488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021659999999999998,
+      "loss": 3.1357,
+      "step": 4593
+    },
+    {
+      "epoch": 7.3504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021645,
+      "loss": 2.927,
+      "step": 4594
+    },
+    {
+      "epoch": 7.352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021629999999999997,
+      "loss": 2.7341,
+      "step": 4595
+    },
+    {
+      "epoch": 7.3536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021615,
+      "loss": 2.8885,
+      "step": 4596
+    },
+    {
+      "epoch": 7.3552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 2.8116,
+      "step": 4597
+    },
+    {
+      "epoch": 7.3568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021585,
+      "loss": 2.6231,
+      "step": 4598
+    },
+    {
+      "epoch": 7.3584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021569999999999998,
+      "loss": 2.3804,
+      "step": 4599
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021554999999999998,
+      "loss": 2.9015,
+      "step": 4600
+    },
+    {
+      "epoch": 7.3616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021539999999999998,
+      "loss": 2.782,
+      "step": 4601
+    },
+    {
+      "epoch": 7.3632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021525,
+      "loss": 2.5905,
+      "step": 4602
+    },
+    {
+      "epoch": 7.3648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021509999999999997,
+      "loss": 2.6103,
+      "step": 4603
+    },
+    {
+      "epoch": 7.3664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021495,
+      "loss": 2.575,
+      "step": 4604
+    },
+    {
+      "epoch": 7.368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021479999999999996,
+      "loss": 3.0087,
+      "step": 4605
+    },
+    {
+      "epoch": 7.3696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021464999999999999,
+      "loss": 2.3155,
+      "step": 4606
+    },
+    {
+      "epoch": 7.3712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021449999999999998,
+      "loss": 2.8529,
+      "step": 4607
+    },
+    {
+      "epoch": 7.3728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021434999999999998,
+      "loss": 2.7225,
+      "step": 4608
+    },
+    {
+      "epoch": 7.3744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021419999999999998,
+      "loss": 2.8051,
+      "step": 4609
+    },
+    {
+      "epoch": 7.376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021405,
+      "loss": 2.5828,
+      "step": 4610
+    },
+    {
+      "epoch": 7.3776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021389999999999997,
+      "loss": 3.0833,
+      "step": 4611
+    },
+    {
+      "epoch": 7.3792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021375,
+      "loss": 2.8253,
+      "step": 4612
+    },
+    {
+      "epoch": 7.3808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021359999999999996,
+      "loss": 2.7968,
+      "step": 4613
+    },
+    {
+      "epoch": 7.3824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021344999999999998,
+      "loss": 2.8949,
+      "step": 4614
+    },
+    {
+      "epoch": 7.384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021329999999999998,
+      "loss": 2.5806,
+      "step": 4615
+    },
+    {
+      "epoch": 7.3856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021314999999999998,
+      "loss": 3.7758,
+      "step": 4616
+    },
+    {
+      "epoch": 7.3872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 2.7733,
+      "step": 4617
+    },
+    {
+      "epoch": 7.3888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021285,
+      "loss": 3.2324,
+      "step": 4618
+    },
+    {
+      "epoch": 7.3904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021269999999999997,
+      "loss": 2.5837,
+      "step": 4619
+    },
+    {
+      "epoch": 7.392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021255,
+      "loss": 2.8423,
+      "step": 4620
+    },
+    {
+      "epoch": 7.3936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021239999999999996,
+      "loss": 2.7471,
+      "step": 4621
+    },
+    {
+      "epoch": 7.3952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021224999999999998,
+      "loss": 2.7898,
+      "step": 4622
+    },
+    {
+      "epoch": 7.3968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021209999999999998,
+      "loss": 3.2034,
+      "step": 4623
+    },
+    {
+      "epoch": 7.3984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021194999999999997,
+      "loss": 3.224,
+      "step": 4624
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021179999999999997,
+      "loss": 3.9561,
+      "step": 4625
+    },
+    {
+      "epoch": 7.4016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021165,
+      "loss": 4.0521,
+      "step": 4626
+    },
+    {
+      "epoch": 7.4032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021149999999999996,
+      "loss": 4.0914,
+      "step": 4627
+    },
+    {
+      "epoch": 7.4048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021135,
+      "loss": 3.4127,
+      "step": 4628
+    },
+    {
+      "epoch": 7.4064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021119999999999996,
+      "loss": 3.3387,
+      "step": 4629
+    },
+    {
+      "epoch": 7.408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021104999999999998,
+      "loss": 3.7805,
+      "step": 4630
+    },
+    {
+      "epoch": 7.4096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021089999999999998,
+      "loss": 3.903,
+      "step": 4631
+    },
+    {
+      "epoch": 7.4112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021074999999999997,
+      "loss": 3.3901,
+      "step": 4632
+    },
+    {
+      "epoch": 7.4128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021059999999999997,
+      "loss": 2.9554,
+      "step": 4633
+    },
+    {
+      "epoch": 7.4144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021045,
+      "loss": 3.3087,
+      "step": 4634
+    },
+    {
+      "epoch": 7.416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021029999999999996,
+      "loss": 2.9068,
+      "step": 4635
+    },
+    {
+      "epoch": 7.4176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00021014999999999999,
+      "loss": 2.7432,
+      "step": 4636
+    },
+    {
+      "epoch": 7.4192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 2.867,
+      "step": 4637
+    },
+    {
+      "epoch": 7.4208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020984999999999998,
+      "loss": 3.2016,
+      "step": 4638
+    },
+    {
+      "epoch": 7.4224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020969999999999997,
+      "loss": 3.1958,
+      "step": 4639
+    },
+    {
+      "epoch": 7.424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020955,
+      "loss": 3.1179,
+      "step": 4640
+    },
+    {
+      "epoch": 7.4256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020939999999999997,
+      "loss": 2.5467,
+      "step": 4641
+    },
+    {
+      "epoch": 7.4272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020925,
+      "loss": 3.0303,
+      "step": 4642
+    },
+    {
+      "epoch": 7.4288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020909999999999996,
+      "loss": 2.5878,
+      "step": 4643
+    },
+    {
+      "epoch": 7.4304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020894999999999998,
+      "loss": 3.0959,
+      "step": 4644
+    },
+    {
+      "epoch": 7.432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020879999999999998,
+      "loss": 2.7835,
+      "step": 4645
+    },
+    {
+      "epoch": 7.4336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020864999999999998,
+      "loss": 3.2442,
+      "step": 4646
+    },
+    {
+      "epoch": 7.4352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020849999999999997,
+      "loss": 2.6535,
+      "step": 4647
+    },
+    {
+      "epoch": 7.4368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020835,
+      "loss": 2.9818,
+      "step": 4648
+    },
+    {
+      "epoch": 7.4384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020819999999999996,
+      "loss": 2.9469,
+      "step": 4649
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020805,
+      "loss": 3.2529,
+      "step": 4650
+    },
+    {
+      "epoch": 7.4416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020789999999999996,
+      "loss": 3.6801,
+      "step": 4651
+    },
+    {
+      "epoch": 7.4432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020774999999999998,
+      "loss": 2.682,
+      "step": 4652
+    },
+    {
+      "epoch": 7.4448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020759999999999998,
+      "loss": 2.9149,
+      "step": 4653
+    },
+    {
+      "epoch": 7.4464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020744999999999997,
+      "loss": 3.1497,
+      "step": 4654
+    },
+    {
+      "epoch": 7.448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020729999999999997,
+      "loss": 2.9631,
+      "step": 4655
+    },
+    {
+      "epoch": 7.4496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020715,
+      "loss": 2.767,
+      "step": 4656
+    },
+    {
+      "epoch": 7.4512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020699999999999996,
+      "loss": 3.045,
+      "step": 4657
+    },
+    {
+      "epoch": 7.4528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020684999999999999,
+      "loss": 2.5517,
+      "step": 4658
+    },
+    {
+      "epoch": 7.4544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020669999999999996,
+      "loss": 3.8156,
+      "step": 4659
+    },
+    {
+      "epoch": 7.456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020654999999999998,
+      "loss": 3.2374,
+      "step": 4660
+    },
+    {
+      "epoch": 7.4576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020639999999999998,
+      "loss": 2.8912,
+      "step": 4661
+    },
+    {
+      "epoch": 7.4592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020624999999999997,
+      "loss": 2.6571,
+      "step": 4662
+    },
+    {
+      "epoch": 7.4608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002061,
+      "loss": 2.9825,
+      "step": 4663
+    },
+    {
+      "epoch": 7.4624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020595,
+      "loss": 3.2419,
+      "step": 4664
+    },
+    {
+      "epoch": 7.464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002058,
+      "loss": 2.3981,
+      "step": 4665
+    },
+    {
+      "epoch": 7.4656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020564999999999998,
+      "loss": 2.7962,
+      "step": 4666
+    },
+    {
+      "epoch": 7.4672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002055,
+      "loss": 3.2975,
+      "step": 4667
+    },
+    {
+      "epoch": 7.4688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020534999999999998,
+      "loss": 3.0376,
+      "step": 4668
+    },
+    {
+      "epoch": 7.4704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002052,
+      "loss": 3.1567,
+      "step": 4669
+    },
+    {
+      "epoch": 7.4719999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020504999999999997,
+      "loss": 3.283,
+      "step": 4670
+    },
+    {
+      "epoch": 7.4736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002049,
+      "loss": 2.2496,
+      "step": 4671
+    },
+    {
+      "epoch": 7.4752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020475,
+      "loss": 2.6953,
+      "step": 4672
+    },
+    {
+      "epoch": 7.4768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020459999999999999,
+      "loss": 3.5792,
+      "step": 4673
+    },
+    {
+      "epoch": 7.4784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020444999999999998,
+      "loss": 3.2485,
+      "step": 4674
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002043,
+      "loss": 3.1137,
+      "step": 4675
+    },
+    {
+      "epoch": 7.4816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020414999999999997,
+      "loss": 4.1653,
+      "step": 4676
+    },
+    {
+      "epoch": 7.4832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000204,
+      "loss": 4.5673,
+      "step": 4677
+    },
+    {
+      "epoch": 7.4848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020384999999999997,
+      "loss": 3.773,
+      "step": 4678
+    },
+    {
+      "epoch": 7.4864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002037,
+      "loss": 3.8248,
+      "step": 4679
+    },
+    {
+      "epoch": 7.4879999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020355,
+      "loss": 3.3055,
+      "step": 4680
+    },
+    {
+      "epoch": 7.4896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020339999999999998,
+      "loss": 3.5717,
+      "step": 4681
+    },
+    {
+      "epoch": 7.4912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020324999999999998,
+      "loss": 3.0292,
+      "step": 4682
+    },
+    {
+      "epoch": 7.4928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002031,
+      "loss": 4.1389,
+      "step": 4683
+    },
+    {
+      "epoch": 7.4944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020294999999999997,
+      "loss": 2.9023,
+      "step": 4684
+    },
+    {
+      "epoch": 7.496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002028,
+      "loss": 2.8841,
+      "step": 4685
+    },
+    {
+      "epoch": 7.4976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020264999999999997,
+      "loss": 3.5379,
+      "step": 4686
+    },
+    {
+      "epoch": 7.4992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002025,
+      "loss": 3.1102,
+      "step": 4687
+    },
+    {
+      "epoch": 7.5008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020234999999999999,
+      "loss": 2.3866,
+      "step": 4688
+    },
+    {
+      "epoch": 7.5024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002022,
+      "loss": 3.1948,
+      "step": 4689
+    },
+    {
+      "epoch": 7.504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020204999999999998,
+      "loss": 2.9193,
+      "step": 4690
+    },
+    {
+      "epoch": 7.5056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002019,
+      "loss": 2.6375,
+      "step": 4691
+    },
+    {
+      "epoch": 7.5072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020174999999999997,
+      "loss": 2.7399,
+      "step": 4692
+    },
+    {
+      "epoch": 7.5088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002016,
+      "loss": 2.7869,
+      "step": 4693
+    },
+    {
+      "epoch": 7.5104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020145,
+      "loss": 2.7297,
+      "step": 4694
+    },
+    {
+      "epoch": 7.5120000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002013,
+      "loss": 3.0658,
+      "step": 4695
+    },
+    {
+      "epoch": 7.5136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020114999999999998,
+      "loss": 2.5824,
+      "step": 4696
+    },
+    {
+      "epoch": 7.5152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000201,
+      "loss": 3.1716,
+      "step": 4697
+    },
+    {
+      "epoch": 7.5168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020084999999999998,
+      "loss": 3.4955,
+      "step": 4698
+    },
+    {
+      "epoch": 7.5184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002007,
+      "loss": 2.6723,
+      "step": 4699
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020054999999999997,
+      "loss": 3.1623,
+      "step": 4700
+    },
+    {
+      "epoch": 7.5216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0002004,
+      "loss": 2.3293,
+      "step": 4701
+    },
+    {
+      "epoch": 7.5232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020025,
+      "loss": 3.2434,
+      "step": 4702
+    },
+    {
+      "epoch": 7.5248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00020009999999999998,
+      "loss": 3.1604,
+      "step": 4703
+    },
+    {
+      "epoch": 7.5264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019994999999999998,
+      "loss": 2.4726,
+      "step": 4704
+    },
+    {
+      "epoch": 7.5280000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001998,
+      "loss": 3.4306,
+      "step": 4705
+    },
+    {
+      "epoch": 7.5296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019964999999999997,
+      "loss": 2.6647,
+      "step": 4706
+    },
+    {
+      "epoch": 7.5312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001995,
+      "loss": 2.991,
+      "step": 4707
+    },
+    {
+      "epoch": 7.5328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019934999999999997,
+      "loss": 3.0528,
+      "step": 4708
+    },
+    {
+      "epoch": 7.5344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001992,
+      "loss": 3.1222,
+      "step": 4709
+    },
+    {
+      "epoch": 7.536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019905,
+      "loss": 3.3261,
+      "step": 4710
+    },
+    {
+      "epoch": 7.5376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019889999999999998,
+      "loss": 2.8421,
+      "step": 4711
+    },
+    {
+      "epoch": 7.5392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019874999999999998,
+      "loss": 2.7211,
+      "step": 4712
+    },
+    {
+      "epoch": 7.5408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001986,
+      "loss": 2.951,
+      "step": 4713
+    },
+    {
+      "epoch": 7.5424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019844999999999997,
+      "loss": 3.3668,
+      "step": 4714
+    },
+    {
+      "epoch": 7.5440000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001983,
+      "loss": 2.5635,
+      "step": 4715
+    },
+    {
+      "epoch": 7.5456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019814999999999996,
+      "loss": 3.2249,
+      "step": 4716
+    },
+    {
+      "epoch": 7.5472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000198,
+      "loss": 2.7145,
+      "step": 4717
+    },
+    {
+      "epoch": 7.5488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019784999999999998,
+      "loss": 2.9497,
+      "step": 4718
+    },
+    {
+      "epoch": 7.5504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019769999999999998,
+      "loss": 3.2384,
+      "step": 4719
+    },
+    {
+      "epoch": 7.552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019754999999999998,
+      "loss": 2.9618,
+      "step": 4720
+    },
+    {
+      "epoch": 7.5536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001974,
+      "loss": 3.3049,
+      "step": 4721
+    },
+    {
+      "epoch": 7.5552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019724999999999997,
+      "loss": 3.2177,
+      "step": 4722
+    },
+    {
+      "epoch": 7.5568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001971,
+      "loss": 3.6352,
+      "step": 4723
+    },
+    {
+      "epoch": 7.5584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019694999999999996,
+      "loss": 3.0243,
+      "step": 4724
+    },
+    {
+      "epoch": 7.5600000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019679999999999999,
+      "loss": 4.1826,
+      "step": 4725
+    },
+    {
+      "epoch": 7.5616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019664999999999998,
+      "loss": 4.4972,
+      "step": 4726
+    },
+    {
+      "epoch": 7.5632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019649999999999998,
+      "loss": 4.1343,
+      "step": 4727
+    },
+    {
+      "epoch": 7.5648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019634999999999998,
+      "loss": 3.8212,
+      "step": 4728
+    },
+    {
+      "epoch": 7.5664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001962,
+      "loss": 3.3201,
+      "step": 4729
+    },
+    {
+      "epoch": 7.568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019604999999999997,
+      "loss": 3.2459,
+      "step": 4730
+    },
+    {
+      "epoch": 7.5696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001959,
+      "loss": 3.7524,
+      "step": 4731
+    },
+    {
+      "epoch": 7.5712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019574999999999996,
+      "loss": 3.0224,
+      "step": 4732
+    },
+    {
+      "epoch": 7.5728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019559999999999998,
+      "loss": 3.2785,
+      "step": 4733
+    },
+    {
+      "epoch": 7.5744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019544999999999998,
+      "loss": 3.0942,
+      "step": 4734
+    },
+    {
+      "epoch": 7.576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019529999999999998,
+      "loss": 3.1707,
+      "step": 4735
+    },
+    {
+      "epoch": 7.5776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019514999999999997,
+      "loss": 2.686,
+      "step": 4736
+    },
+    {
+      "epoch": 7.5792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000195,
+      "loss": 2.6039,
+      "step": 4737
+    },
+    {
+      "epoch": 7.5808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019484999999999997,
+      "loss": 3.097,
+      "step": 4738
+    },
+    {
+      "epoch": 7.5824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001947,
+      "loss": 2.4248,
+      "step": 4739
+    },
+    {
+      "epoch": 7.584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019454999999999999,
+      "loss": 2.9075,
+      "step": 4740
+    },
+    {
+      "epoch": 7.5856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019439999999999998,
+      "loss": 2.7502,
+      "step": 4741
+    },
+    {
+      "epoch": 7.5872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019424999999999998,
+      "loss": 3.0797,
+      "step": 4742
+    },
+    {
+      "epoch": 7.5888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001941,
+      "loss": 3.082,
+      "step": 4743
+    },
+    {
+      "epoch": 7.5904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019394999999999997,
+      "loss": 3.4958,
+      "step": 4744
+    },
+    {
+      "epoch": 7.592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001938,
+      "loss": 2.8035,
+      "step": 4745
+    },
+    {
+      "epoch": 7.5936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019364999999999996,
+      "loss": 2.5721,
+      "step": 4746
+    },
+    {
+      "epoch": 7.5952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001935,
+      "loss": 2.7212,
+      "step": 4747
+    },
+    {
+      "epoch": 7.5968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019334999999999998,
+      "loss": 3.8946,
+      "step": 4748
+    },
+    {
+      "epoch": 7.5984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019319999999999998,
+      "loss": 2.6282,
+      "step": 4749
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019304999999999998,
+      "loss": 2.9679,
+      "step": 4750
+    },
+    {
+      "epoch": 7.6016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001929,
+      "loss": 2.5958,
+      "step": 4751
+    },
+    {
+      "epoch": 7.6032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019274999999999997,
+      "loss": 3.128,
+      "step": 4752
+    },
+    {
+      "epoch": 7.6048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001926,
+      "loss": 2.3277,
+      "step": 4753
+    },
+    {
+      "epoch": 7.6064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019244999999999996,
+      "loss": 2.5896,
+      "step": 4754
+    },
+    {
+      "epoch": 7.608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019229999999999999,
+      "loss": 2.8047,
+      "step": 4755
+    },
+    {
+      "epoch": 7.6096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019214999999999998,
+      "loss": 3.0452,
+      "step": 4756
+    },
+    {
+      "epoch": 7.6112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 2.8145,
+      "step": 4757
+    },
+    {
+      "epoch": 7.6128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019184999999999997,
+      "loss": 2.6334,
+      "step": 4758
+    },
+    {
+      "epoch": 7.6144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001917,
+      "loss": 2.5752,
+      "step": 4759
+    },
+    {
+      "epoch": 7.616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019154999999999997,
+      "loss": 3.1652,
+      "step": 4760
+    },
+    {
+      "epoch": 7.6176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001914,
+      "loss": 2.452,
+      "step": 4761
+    },
+    {
+      "epoch": 7.6192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019124999999999996,
+      "loss": 2.4447,
+      "step": 4762
+    },
+    {
+      "epoch": 7.6208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019109999999999998,
+      "loss": 2.3181,
+      "step": 4763
+    },
+    {
+      "epoch": 7.6224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019094999999999998,
+      "loss": 2.6932,
+      "step": 4764
+    },
+    {
+      "epoch": 7.624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019079999999999998,
+      "loss": 3.6925,
+      "step": 4765
+    },
+    {
+      "epoch": 7.6256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019064999999999997,
+      "loss": 4.25,
+      "step": 4766
+    },
+    {
+      "epoch": 7.6272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001905,
+      "loss": 3.0461,
+      "step": 4767
+    },
+    {
+      "epoch": 7.6288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019034999999999996,
+      "loss": 3.0939,
+      "step": 4768
+    },
+    {
+      "epoch": 7.6304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001902,
+      "loss": 2.8625,
+      "step": 4769
+    },
+    {
+      "epoch": 7.632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019004999999999996,
+      "loss": 3.7221,
+      "step": 4770
+    },
+    {
+      "epoch": 7.6336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018989999999999998,
+      "loss": 2.636,
+      "step": 4771
+    },
+    {
+      "epoch": 7.6352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018974999999999998,
+      "loss": 3.3965,
+      "step": 4772
+    },
+    {
+      "epoch": 7.6368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018959999999999997,
+      "loss": 2.7437,
+      "step": 4773
+    },
+    {
+      "epoch": 7.6384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018944999999999997,
+      "loss": 2.915,
+      "step": 4774
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": NaN,
+      "learning_rate": 0.00018944999999999997,
+      "loss": 2.9823,
+      "step": 4775
+    },
+    {
+      "epoch": 7.6416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001893,
+      "loss": 3.0832,
+      "step": 4776
+    },
+    {
+      "epoch": 7.6432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018914999999999996,
+      "loss": 3.059,
+      "step": 4777
+    },
+    {
+      "epoch": 7.6448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 2.8947,
+      "step": 4778
+    },
+    {
+      "epoch": 7.6464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018884999999999996,
+      "loss": 3.123,
+      "step": 4779
+    },
+    {
+      "epoch": 7.648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018869999999999998,
+      "loss": 3.7362,
+      "step": 4780
+    },
+    {
+      "epoch": 7.6495999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018854999999999998,
+      "loss": 3.4403,
+      "step": 4781
+    },
+    {
+      "epoch": 7.6512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018839999999999997,
+      "loss": 2.7382,
+      "step": 4782
+    },
+    {
+      "epoch": 7.6528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018824999999999997,
+      "loss": 3.2175,
+      "step": 4783
+    },
+    {
+      "epoch": 7.6544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001881,
+      "loss": 3.1441,
+      "step": 4784
+    },
+    {
+      "epoch": 7.656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018794999999999996,
+      "loss": 3.2875,
+      "step": 4785
+    },
+    {
+      "epoch": 7.6576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018779999999999998,
+      "loss": 2.4975,
+      "step": 4786
+    },
+    {
+      "epoch": 7.6592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018764999999999998,
+      "loss": 3.5703,
+      "step": 4787
+    },
+    {
+      "epoch": 7.6608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 2.8902,
+      "step": 4788
+    },
+    {
+      "epoch": 7.6624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018735,
+      "loss": 3.2099,
+      "step": 4789
+    },
+    {
+      "epoch": 7.664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001872,
+      "loss": 2.9238,
+      "step": 4790
+    },
+    {
+      "epoch": 7.6655999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018705,
+      "loss": 3.0309,
+      "step": 4791
+    },
+    {
+      "epoch": 7.6672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001869,
+      "loss": 3.3599,
+      "step": 4792
+    },
+    {
+      "epoch": 7.6688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018675,
+      "loss": 2.6549,
+      "step": 4793
+    },
+    {
+      "epoch": 7.6704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018659999999999998,
+      "loss": 3.019,
+      "step": 4794
+    },
+    {
+      "epoch": 7.672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018645,
+      "loss": 2.5279,
+      "step": 4795
+    },
+    {
+      "epoch": 7.6736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018629999999999997,
+      "loss": 2.6226,
+      "step": 4796
+    },
+    {
+      "epoch": 7.6752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018615,
+      "loss": 3.0384,
+      "step": 4797
+    },
+    {
+      "epoch": 7.6768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000186,
+      "loss": 2.7494,
+      "step": 4798
+    },
+    {
+      "epoch": 7.6784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018585,
+      "loss": 2.9542,
+      "step": 4799
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001857,
+      "loss": 2.6497,
+      "step": 4800
+    },
+    {
+      "epoch": 7.6815999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018555,
+      "loss": 2.7592,
+      "step": 4801
+    },
+    {
+      "epoch": 7.6832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018539999999999998,
+      "loss": 2.4912,
+      "step": 4802
+    },
+    {
+      "epoch": 7.6848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018525,
+      "loss": 2.8979,
+      "step": 4803
+    },
+    {
+      "epoch": 7.6864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018509999999999997,
+      "loss": 3.2736,
+      "step": 4804
+    },
+    {
+      "epoch": 7.688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018495,
+      "loss": 3.156,
+      "step": 4805
+    },
+    {
+      "epoch": 7.6896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001848,
+      "loss": 2.873,
+      "step": 4806
+    },
+    {
+      "epoch": 7.6912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018465,
+      "loss": 2.6056,
+      "step": 4807
+    },
+    {
+      "epoch": 7.6928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018449999999999999,
+      "loss": 2.2569,
+      "step": 4808
+    },
+    {
+      "epoch": 7.6944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018435,
+      "loss": 3.3233,
+      "step": 4809
+    },
+    {
+      "epoch": 7.696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018419999999999998,
+      "loss": 2.7609,
+      "step": 4810
+    },
+    {
+      "epoch": 7.6975999999999996,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018405,
+      "loss": 2.8943,
+      "step": 4811
+    },
+    {
+      "epoch": 7.6992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018389999999999997,
+      "loss": 2.8554,
+      "step": 4812
+    },
+    {
+      "epoch": 7.7008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018375,
+      "loss": 3.2062,
+      "step": 4813
+    },
+    {
+      "epoch": 7.7024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001836,
+      "loss": 3.4082,
+      "step": 4814
+    },
+    {
+      "epoch": 7.704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018345,
+      "loss": 2.3895,
+      "step": 4815
+    },
+    {
+      "epoch": 7.7056000000000004,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018329999999999998,
+      "loss": 3.1006,
+      "step": 4816
+    },
+    {
+      "epoch": 7.7072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018315,
+      "loss": 2.6601,
+      "step": 4817
+    },
+    {
+      "epoch": 7.7088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018299999999999998,
+      "loss": 3.321,
+      "step": 4818
+    },
+    {
+      "epoch": 7.7104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018285,
+      "loss": 3.0802,
+      "step": 4819
+    },
+    {
+      "epoch": 7.712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018269999999999997,
+      "loss": 2.8503,
+      "step": 4820
+    },
+    {
+      "epoch": 7.7136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018255,
+      "loss": 2.7178,
+      "step": 4821
+    },
+    {
+      "epoch": 7.7152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001824,
+      "loss": 3.6901,
+      "step": 4822
+    },
+    {
+      "epoch": 7.7168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018224999999999998,
+      "loss": 3.2879,
+      "step": 4823
+    },
+    {
+      "epoch": 7.7184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018209999999999998,
+      "loss": 2.8951,
+      "step": 4824
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018195,
+      "loss": 3.6328,
+      "step": 4825
+    },
+    {
+      "epoch": 7.7216000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018179999999999997,
+      "loss": 4.4299,
+      "step": 4826
+    },
+    {
+      "epoch": 7.7232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018165,
+      "loss": 4.0601,
+      "step": 4827
+    },
+    {
+      "epoch": 7.7248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018149999999999997,
+      "loss": 3.0605,
+      "step": 4828
+    },
+    {
+      "epoch": 7.7264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018135,
+      "loss": 3.3661,
+      "step": 4829
+    },
+    {
+      "epoch": 7.728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018119999999999999,
+      "loss": 3.6533,
+      "step": 4830
+    },
+    {
+      "epoch": 7.7296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018104999999999998,
+      "loss": 4.6602,
+      "step": 4831
+    },
+    {
+      "epoch": 7.7312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018089999999999998,
+      "loss": 3.2386,
+      "step": 4832
+    },
+    {
+      "epoch": 7.7328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018075,
+      "loss": 3.4873,
+      "step": 4833
+    },
+    {
+      "epoch": 7.7344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018059999999999997,
+      "loss": 3.9829,
+      "step": 4834
+    },
+    {
+      "epoch": 7.736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018045,
+      "loss": 4.0437,
+      "step": 4835
+    },
+    {
+      "epoch": 7.7376000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018029999999999996,
+      "loss": 2.6534,
+      "step": 4836
+    },
+    {
+      "epoch": 7.7392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00018015,
+      "loss": 3.3255,
+      "step": 4837
+    },
+    {
+      "epoch": 7.7408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 2.3529,
+      "step": 4838
+    },
+    {
+      "epoch": 7.7424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017984999999999998,
+      "loss": 2.8427,
+      "step": 4839
+    },
+    {
+      "epoch": 7.744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017969999999999998,
+      "loss": 3.0223,
+      "step": 4840
+    },
+    {
+      "epoch": 7.7456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017955,
+      "loss": 2.8049,
+      "step": 4841
+    },
+    {
+      "epoch": 7.7472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017939999999999997,
+      "loss": 2.7341,
+      "step": 4842
+    },
+    {
+      "epoch": 7.7488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017925,
+      "loss": 2.4303,
+      "step": 4843
+    },
+    {
+      "epoch": 7.7504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001791,
+      "loss": 3.6557,
+      "step": 4844
+    },
+    {
+      "epoch": 7.752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017894999999999999,
+      "loss": 2.5856,
+      "step": 4845
+    },
+    {
+      "epoch": 7.7536000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017879999999999998,
+      "loss": 2.7199,
+      "step": 4846
+    },
+    {
+      "epoch": 7.7552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017865,
+      "loss": 2.2769,
+      "step": 4847
+    },
+    {
+      "epoch": 7.7568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017849999999999997,
+      "loss": 3.1895,
+      "step": 4848
+    },
+    {
+      "epoch": 7.7584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017835,
+      "loss": 3.0962,
+      "step": 4849
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017819999999999997,
+      "loss": 2.5096,
+      "step": 4850
+    },
+    {
+      "epoch": 7.7616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017805,
+      "loss": 3.0981,
+      "step": 4851
+    },
+    {
+      "epoch": 7.7632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001779,
+      "loss": 2.7704,
+      "step": 4852
+    },
+    {
+      "epoch": 7.7648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017774999999999998,
+      "loss": 2.6671,
+      "step": 4853
+    },
+    {
+      "epoch": 7.7664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017759999999999998,
+      "loss": 3.685,
+      "step": 4854
+    },
+    {
+      "epoch": 7.768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017745,
+      "loss": 2.9276,
+      "step": 4855
+    },
+    {
+      "epoch": 7.7696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017729999999999997,
+      "loss": 2.6199,
+      "step": 4856
+    },
+    {
+      "epoch": 7.7712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017715,
+      "loss": 3.2054,
+      "step": 4857
+    },
+    {
+      "epoch": 7.7728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017699999999999997,
+      "loss": 2.7798,
+      "step": 4858
+    },
+    {
+      "epoch": 7.7744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017685,
+      "loss": 2.1486,
+      "step": 4859
+    },
+    {
+      "epoch": 7.776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017669999999999999,
+      "loss": 2.8584,
+      "step": 4860
+    },
+    {
+      "epoch": 7.7776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017654999999999998,
+      "loss": 2.9491,
+      "step": 4861
+    },
+    {
+      "epoch": 7.7792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017639999999999998,
+      "loss": 3.88,
+      "step": 4862
+    },
+    {
+      "epoch": 7.7808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017625,
+      "loss": 3.6535,
+      "step": 4863
+    },
+    {
+      "epoch": 7.7824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017609999999999997,
+      "loss": 3.1241,
+      "step": 4864
+    },
+    {
+      "epoch": 7.784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017595,
+      "loss": 3.2194,
+      "step": 4865
+    },
+    {
+      "epoch": 7.7856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017579999999999996,
+      "loss": 3.383,
+      "step": 4866
+    },
+    {
+      "epoch": 7.7872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017565,
+      "loss": 3.2972,
+      "step": 4867
+    },
+    {
+      "epoch": 7.7888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017549999999999998,
+      "loss": 2.985,
+      "step": 4868
+    },
+    {
+      "epoch": 7.7904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017534999999999998,
+      "loss": 2.5855,
+      "step": 4869
+    },
+    {
+      "epoch": 7.792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017519999999999998,
+      "loss": 3.1634,
+      "step": 4870
+    },
+    {
+      "epoch": 7.7936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017505,
+      "loss": 3.1311,
+      "step": 4871
+    },
+    {
+      "epoch": 7.7952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017489999999999997,
+      "loss": 2.5284,
+      "step": 4872
+    },
+    {
+      "epoch": 7.7968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017475,
+      "loss": 3.1961,
+      "step": 4873
+    },
+    {
+      "epoch": 7.7984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017459999999999996,
+      "loss": 3.5777,
+      "step": 4874
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017444999999999998,
+      "loss": 3.6377,
+      "step": 4875
+    },
+    {
+      "epoch": 7.8016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017429999999999998,
+      "loss": 4.5462,
+      "step": 4876
+    },
+    {
+      "epoch": 7.8032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017414999999999998,
+      "loss": 5.1288,
+      "step": 4877
+    },
+    {
+      "epoch": 7.8048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 3.9374,
+      "step": 4878
+    },
+    {
+      "epoch": 7.8064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017385,
+      "loss": 3.9448,
+      "step": 4879
+    },
+    {
+      "epoch": 7.808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017369999999999997,
+      "loss": 3.4438,
+      "step": 4880
+    },
+    {
+      "epoch": 7.8096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017355,
+      "loss": 3.3757,
+      "step": 4881
+    },
+    {
+      "epoch": 7.8112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017339999999999996,
+      "loss": 3.2291,
+      "step": 4882
+    },
+    {
+      "epoch": 7.8128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017324999999999998,
+      "loss": 3.4574,
+      "step": 4883
+    },
+    {
+      "epoch": 7.8144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017309999999999998,
+      "loss": 3.1308,
+      "step": 4884
+    },
+    {
+      "epoch": 7.816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017294999999999998,
+      "loss": 3.5847,
+      "step": 4885
+    },
+    {
+      "epoch": 7.8176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017279999999999997,
+      "loss": 3.464,
+      "step": 4886
+    },
+    {
+      "epoch": 7.8192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017265,
+      "loss": 2.862,
+      "step": 4887
+    },
+    {
+      "epoch": 7.8208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017249999999999996,
+      "loss": 2.6696,
+      "step": 4888
+    },
+    {
+      "epoch": 7.8224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017235,
+      "loss": 3.0537,
+      "step": 4889
+    },
+    {
+      "epoch": 7.824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017219999999999998,
+      "loss": 3.1748,
+      "step": 4890
+    },
+    {
+      "epoch": 7.8256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017204999999999998,
+      "loss": 2.8772,
+      "step": 4891
+    },
+    {
+      "epoch": 7.8272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017189999999999998,
+      "loss": 2.8474,
+      "step": 4892
+    },
+    {
+      "epoch": 7.8288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017175,
+      "loss": 2.9608,
+      "step": 4893
+    },
+    {
+      "epoch": 7.8304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017159999999999997,
+      "loss": 3.2269,
+      "step": 4894
+    },
+    {
+      "epoch": 7.832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017145,
+      "loss": 2.2713,
+      "step": 4895
+    },
+    {
+      "epoch": 7.8336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017129999999999996,
+      "loss": 2.3742,
+      "step": 4896
+    },
+    {
+      "epoch": 7.8352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017114999999999999,
+      "loss": 2.4596,
+      "step": 4897
+    },
+    {
+      "epoch": 7.8368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017099999999999998,
+      "loss": 4.0418,
+      "step": 4898
+    },
+    {
+      "epoch": 7.8384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017084999999999998,
+      "loss": 2.816,
+      "step": 4899
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017069999999999998,
+      "loss": 3.6281,
+      "step": 4900
+    },
+    {
+      "epoch": 7.8416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017055,
+      "loss": 2.9427,
+      "step": 4901
+    },
+    {
+      "epoch": 7.8431999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017039999999999997,
+      "loss": 3.0255,
+      "step": 4902
+    },
+    {
+      "epoch": 7.8448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017025,
+      "loss": 2.7146,
+      "step": 4903
+    },
+    {
+      "epoch": 7.8464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017009999999999996,
+      "loss": 2.7954,
+      "step": 4904
+    },
+    {
+      "epoch": 7.848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016994999999999998,
+      "loss": 2.593,
+      "step": 4905
+    },
+    {
+      "epoch": 7.8496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016979999999999998,
+      "loss": 2.5321,
+      "step": 4906
+    },
+    {
+      "epoch": 7.8512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016964999999999998,
+      "loss": 2.776,
+      "step": 4907
+    },
+    {
+      "epoch": 7.8528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016949999999999997,
+      "loss": 2.554,
+      "step": 4908
+    },
+    {
+      "epoch": 7.8544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016935,
+      "loss": 2.8608,
+      "step": 4909
+    },
+    {
+      "epoch": 7.856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016919999999999997,
+      "loss": 3.0325,
+      "step": 4910
+    },
+    {
+      "epoch": 7.8576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016905,
+      "loss": 2.5525,
+      "step": 4911
+    },
+    {
+      "epoch": 7.8591999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016889999999999996,
+      "loss": 2.5041,
+      "step": 4912
+    },
+    {
+      "epoch": 7.8608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016874999999999998,
+      "loss": 2.4444,
+      "step": 4913
+    },
+    {
+      "epoch": 7.8624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001686,
+      "loss": 2.978,
+      "step": 4914
+    },
+    {
+      "epoch": 7.864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016844999999999997,
+      "loss": 2.5528,
+      "step": 4915
+    },
+    {
+      "epoch": 7.8656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001683,
+      "loss": 3.7146,
+      "step": 4916
+    },
+    {
+      "epoch": 7.8672,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001683,
+      "loss": 2.6056,
+      "step": 4917
+    },
+    {
+      "epoch": 7.8688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016815,
+      "loss": 3.3739,
+      "step": 4918
+    },
+    {
+      "epoch": 7.8704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000168,
+      "loss": 3.0553,
+      "step": 4919
+    },
+    {
+      "epoch": 7.872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016785,
+      "loss": 2.3268,
+      "step": 4920
+    },
+    {
+      "epoch": 7.8736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001677,
+      "loss": 2.5411,
+      "step": 4921
+    },
+    {
+      "epoch": 7.8751999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016754999999999998,
+      "loss": 2.4827,
+      "step": 4922
+    },
+    {
+      "epoch": 7.8768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001674,
+      "loss": 3.9451,
+      "step": 4923
+    },
+    {
+      "epoch": 7.8784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016724999999999997,
+      "loss": 2.621,
+      "step": 4924
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": NaN,
+      "learning_rate": 0.00016724999999999997,
+      "loss": 2.7644,
+      "step": 4925
+    },
+    {
+      "epoch": 7.8816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001671,
+      "loss": 4.6617,
+      "step": 4926
+    },
+    {
+      "epoch": 7.8832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016695,
+      "loss": 3.873,
+      "step": 4927
+    },
+    {
+      "epoch": 7.8848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001668,
+      "loss": 3.3357,
+      "step": 4928
+    },
+    {
+      "epoch": 7.8864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016664999999999998,
+      "loss": 4.252,
+      "step": 4929
+    },
+    {
+      "epoch": 7.888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001665,
+      "loss": 3.3949,
+      "step": 4930
+    },
+    {
+      "epoch": 7.8896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016634999999999998,
+      "loss": 2.5891,
+      "step": 4931
+    },
+    {
+      "epoch": 7.8911999999999995,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001662,
+      "loss": 3.537,
+      "step": 4932
+    },
+    {
+      "epoch": 7.8928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016604999999999997,
+      "loss": 3.3417,
+      "step": 4933
+    },
+    {
+      "epoch": 7.8944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001659,
+      "loss": 2.878,
+      "step": 4934
+    },
+    {
+      "epoch": 7.896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016575,
+      "loss": 3.7171,
+      "step": 4935
+    },
+    {
+      "epoch": 7.8976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001656,
+      "loss": 2.6422,
+      "step": 4936
+    },
+    {
+      "epoch": 7.8992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016544999999999998,
+      "loss": 3.0546,
+      "step": 4937
+    },
+    {
+      "epoch": 7.9008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001653,
+      "loss": 2.7797,
+      "step": 4938
+    },
+    {
+      "epoch": 7.9024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016514999999999998,
+      "loss": 2.5424,
+      "step": 4939
+    },
+    {
+      "epoch": 7.904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000165,
+      "loss": 2.8186,
+      "step": 4940
+    },
+    {
+      "epoch": 7.9056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016485,
+      "loss": 2.8034,
+      "step": 4941
+    },
+    {
+      "epoch": 7.9072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001647,
+      "loss": 3.2561,
+      "step": 4942
+    },
+    {
+      "epoch": 7.9088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016455,
+      "loss": 2.6736,
+      "step": 4943
+    },
+    {
+      "epoch": 7.9104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001644,
+      "loss": 2.5695,
+      "step": 4944
+    },
+    {
+      "epoch": 7.912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016424999999999998,
+      "loss": 2.9508,
+      "step": 4945
+    },
+    {
+      "epoch": 7.9136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001641,
+      "loss": 3.2022,
+      "step": 4946
+    },
+    {
+      "epoch": 7.9152000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016394999999999997,
+      "loss": 2.4282,
+      "step": 4947
+    },
+    {
+      "epoch": 7.9168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001638,
+      "loss": 2.4215,
+      "step": 4948
+    },
+    {
+      "epoch": 7.9184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016365,
+      "loss": 3.4899,
+      "step": 4949
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001635,
+      "loss": 2.93,
+      "step": 4950
+    },
+    {
+      "epoch": 7.9216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016334999999999999,
+      "loss": 2.541,
+      "step": 4951
+    },
+    {
+      "epoch": 7.9232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001632,
+      "loss": 2.9613,
+      "step": 4952
+    },
+    {
+      "epoch": 7.9248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016304999999999998,
+      "loss": 3.1947,
+      "step": 4953
+    },
+    {
+      "epoch": 7.9264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001629,
+      "loss": 2.3242,
+      "step": 4954
+    },
+    {
+      "epoch": 7.928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016274999999999997,
+      "loss": 2.7264,
+      "step": 4955
+    },
+    {
+      "epoch": 7.9296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001626,
+      "loss": 3.0834,
+      "step": 4956
+    },
+    {
+      "epoch": 7.9312000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016245,
+      "loss": 2.9288,
+      "step": 4957
+    },
+    {
+      "epoch": 7.9328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001623,
+      "loss": 3.3546,
+      "step": 4958
+    },
+    {
+      "epoch": 7.9344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016214999999999998,
+      "loss": 2.814,
+      "step": 4959
+    },
+    {
+      "epoch": 7.936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000162,
+      "loss": 3.0695,
+      "step": 4960
+    },
+    {
+      "epoch": 7.9376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016184999999999998,
+      "loss": 2.59,
+      "step": 4961
+    },
+    {
+      "epoch": 7.9392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001617,
+      "loss": 3.9635,
+      "step": 4962
+    },
+    {
+      "epoch": 7.9408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016154999999999997,
+      "loss": 2.8278,
+      "step": 4963
+    },
+    {
+      "epoch": 7.9424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001614,
+      "loss": 2.9744,
+      "step": 4964
+    },
+    {
+      "epoch": 7.944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016125,
+      "loss": 2.4744,
+      "step": 4965
+    },
+    {
+      "epoch": 7.9456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016109999999999999,
+      "loss": 2.4728,
+      "step": 4966
+    },
+    {
+      "epoch": 7.9472000000000005,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016094999999999998,
+      "loss": 2.719,
+      "step": 4967
+    },
+    {
+      "epoch": 7.9488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001608,
+      "loss": 2.5483,
+      "step": 4968
+    },
+    {
+      "epoch": 7.9504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016064999999999997,
+      "loss": 3.4515,
+      "step": 4969
+    },
+    {
+      "epoch": 7.952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001605,
+      "loss": 3.9347,
+      "step": 4970
+    },
+    {
+      "epoch": 7.9536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016034999999999997,
+      "loss": 2.8573,
+      "step": 4971
+    },
+    {
+      "epoch": 7.9552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001602,
+      "loss": 3.3291,
+      "step": 4972
+    },
+    {
+      "epoch": 7.9568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016005,
+      "loss": 3.4811,
+      "step": 4973
+    },
+    {
+      "epoch": 7.9584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015989999999999998,
+      "loss": 3.1149,
+      "step": 4974
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015974999999999998,
+      "loss": 3.6624,
+      "step": 4975
+    },
+    {
+      "epoch": 7.9616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001596,
+      "loss": 4.6154,
+      "step": 4976
+    },
+    {
+      "epoch": 7.9632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015944999999999997,
+      "loss": 3.8381,
+      "step": 4977
+    },
+    {
+      "epoch": 7.9648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001593,
+      "loss": 3.4302,
+      "step": 4978
+    },
+    {
+      "epoch": 7.9664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015914999999999997,
+      "loss": 2.838,
+      "step": 4979
+    },
+    {
+      "epoch": 7.968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000159,
+      "loss": 3.2424,
+      "step": 4980
+    },
+    {
+      "epoch": 7.9696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015884999999999999,
+      "loss": 3.268,
+      "step": 4981
+    },
+    {
+      "epoch": 7.9712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015869999999999998,
+      "loss": 2.5788,
+      "step": 4982
+    },
+    {
+      "epoch": 7.9728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015854999999999998,
+      "loss": 3.1136,
+      "step": 4983
+    },
+    {
+      "epoch": 7.9744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001584,
+      "loss": 3.21,
+      "step": 4984
+    },
+    {
+      "epoch": 7.976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015824999999999997,
+      "loss": 3.0565,
+      "step": 4985
+    },
+    {
+      "epoch": 7.9776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001581,
+      "loss": 2.467,
+      "step": 4986
+    },
+    {
+      "epoch": 7.9792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015794999999999996,
+      "loss": 2.4044,
+      "step": 4987
+    },
+    {
+      "epoch": 7.9808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001578,
+      "loss": 2.5883,
+      "step": 4988
+    },
+    {
+      "epoch": 7.9824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015764999999999998,
+      "loss": 2.5788,
+      "step": 4989
+    },
+    {
+      "epoch": 7.984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015749999999999998,
+      "loss": 3.0362,
+      "step": 4990
+    },
+    {
+      "epoch": 7.9856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015734999999999998,
+      "loss": 2.5262,
+      "step": 4991
+    },
+    {
+      "epoch": 7.9872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001572,
+      "loss": 2.9619,
+      "step": 4992
+    },
+    {
+      "epoch": 7.9888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015704999999999997,
+      "loss": 4.0168,
+      "step": 4993
+    },
+    {
+      "epoch": 7.9904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001569,
+      "loss": 2.4526,
+      "step": 4994
+    },
+    {
+      "epoch": 7.992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015675,
+      "loss": 3.016,
+      "step": 4995
+    },
+    {
+      "epoch": 7.9936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015659999999999998,
+      "loss": 2.4731,
+      "step": 4996
+    },
+    {
+      "epoch": 7.9952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015644999999999998,
+      "loss": 2.3664,
+      "step": 4997
+    },
+    {
+      "epoch": 7.9968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001563,
+      "loss": 2.4883,
+      "step": 4998
+    },
+    {
+      "epoch": 7.9984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015614999999999997,
+      "loss": 2.6371,
+      "step": 4999
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000156,
+      "loss": 3.4098,
+      "step": 5000
+    },
+    {
+      "epoch": 8.0,
+      "eval_cer": 0.5195918204647826,
+      "eval_loss": 3.2885050773620605,
+      "eval_runtime": 161.1911,
+      "eval_samples_per_second": 19.455,
+      "eval_steps_per_second": 1.216,
+      "eval_wer": 0.8050150243498083,
+      "step": 5000
+    },
+    {
+      "epoch": 8.0016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015584999999999997,
+      "loss": 4.2925,
+      "step": 5001
+    },
+    {
+      "epoch": 8.0032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001557,
+      "loss": 4.3188,
+      "step": 5002
+    },
+    {
+      "epoch": 8.0048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015554999999999999,
+      "loss": 3.3977,
+      "step": 5003
+    },
+    {
+      "epoch": 8.0064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015539999999999998,
+      "loss": 3.9065,
+      "step": 5004
+    },
+    {
+      "epoch": 8.008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015524999999999998,
+      "loss": 2.8307,
+      "step": 5005
+    },
+    {
+      "epoch": 8.0096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001551,
+      "loss": 3.3355,
+      "step": 5006
+    },
+    {
+      "epoch": 8.0112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015494999999999997,
+      "loss": 3.7054,
+      "step": 5007
+    },
+    {
+      "epoch": 8.0128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001548,
+      "loss": 3.4316,
+      "step": 5008
+    },
+    {
+      "epoch": 8.0144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015464999999999996,
+      "loss": 2.7287,
+      "step": 5009
+    },
+    {
+      "epoch": 8.016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001545,
+      "loss": 2.9146,
+      "step": 5010
+    },
+    {
+      "epoch": 8.0176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015434999999999998,
+      "loss": 3.566,
+      "step": 5011
+    },
+    {
+      "epoch": 8.0192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015419999999999998,
+      "loss": 3.9366,
+      "step": 5012
+    },
+    {
+      "epoch": 8.0208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015404999999999998,
+      "loss": 2.746,
+      "step": 5013
+    },
+    {
+      "epoch": 8.0224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001539,
+      "loss": 2.7101,
+      "step": 5014
+    },
+    {
+      "epoch": 8.024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015374999999999997,
+      "loss": 3.2882,
+      "step": 5015
+    },
+    {
+      "epoch": 8.0256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001536,
+      "loss": 2.9917,
+      "step": 5016
+    },
+    {
+      "epoch": 8.0272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015344999999999996,
+      "loss": 2.8916,
+      "step": 5017
+    },
+    {
+      "epoch": 8.0288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015329999999999999,
+      "loss": 2.4952,
+      "step": 5018
+    },
+    {
+      "epoch": 8.0304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015314999999999998,
+      "loss": 2.9952,
+      "step": 5019
+    },
+    {
+      "epoch": 8.032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015299999999999998,
+      "loss": 2.9704,
+      "step": 5020
+    },
+    {
+      "epoch": 8.0336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015284999999999997,
+      "loss": 2.6999,
+      "step": 5021
+    },
+    {
+      "epoch": 8.0352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001527,
+      "loss": 2.2176,
+      "step": 5022
+    },
+    {
+      "epoch": 8.0368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015254999999999997,
+      "loss": 2.576,
+      "step": 5023
+    },
+    {
+      "epoch": 8.0384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001524,
+      "loss": 2.979,
+      "step": 5024
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015224999999999996,
+      "loss": 2.9045,
+      "step": 5025
+    },
+    {
+      "epoch": 8.0416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015209999999999998,
+      "loss": 2.8305,
+      "step": 5026
+    },
+    {
+      "epoch": 8.0432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015194999999999998,
+      "loss": 2.6715,
+      "step": 5027
+    },
+    {
+      "epoch": 8.0448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015179999999999998,
+      "loss": 2.3777,
+      "step": 5028
+    },
+    {
+      "epoch": 8.0464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015164999999999997,
+      "loss": 2.6145,
+      "step": 5029
+    },
+    {
+      "epoch": 8.048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001515,
+      "loss": 2.9689,
+      "step": 5030
+    },
+    {
+      "epoch": 8.0496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015134999999999997,
+      "loss": 2.6881,
+      "step": 5031
+    },
+    {
+      "epoch": 8.0512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001512,
+      "loss": 2.4856,
+      "step": 5032
+    },
+    {
+      "epoch": 8.0528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015104999999999996,
+      "loss": 2.4703,
+      "step": 5033
+    },
+    {
+      "epoch": 8.0544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015089999999999998,
+      "loss": 3.364,
+      "step": 5034
+    },
+    {
+      "epoch": 8.056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015074999999999998,
+      "loss": 2.7439,
+      "step": 5035
+    },
+    {
+      "epoch": 8.0576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015059999999999997,
+      "loss": 2.6949,
+      "step": 5036
+    },
+    {
+      "epoch": 8.0592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015044999999999997,
+      "loss": 3.5367,
+      "step": 5037
+    },
+    {
+      "epoch": 8.0608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001503,
+      "loss": 2.4536,
+      "step": 5038
+    },
+    {
+      "epoch": 8.0624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015014999999999996,
+      "loss": 2.1861,
+      "step": 5039
+    },
+    {
+      "epoch": 8.064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015,
+      "loss": 3.0478,
+      "step": 5040
+    },
+    {
+      "epoch": 8.0656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014984999999999998,
+      "loss": 2.8823,
+      "step": 5041
+    },
+    {
+      "epoch": 8.0672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014969999999999998,
+      "loss": 3.1218,
+      "step": 5042
+    },
+    {
+      "epoch": 8.0688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014954999999999998,
+      "loss": 2.5482,
+      "step": 5043
+    },
+    {
+      "epoch": 8.0704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001494,
+      "loss": 3.969,
+      "step": 5044
+    },
+    {
+      "epoch": 8.072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014925,
+      "loss": 2.7562,
+      "step": 5045
+    },
+    {
+      "epoch": 8.0736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001491,
+      "loss": 3.4103,
+      "step": 5046
+    },
+    {
+      "epoch": 8.0752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014895,
+      "loss": 4.3819,
+      "step": 5047
+    },
+    {
+      "epoch": 8.0768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014879999999999998,
+      "loss": 2.7046,
+      "step": 5048
+    },
+    {
+      "epoch": 8.0784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014864999999999998,
+      "loss": 4.2045,
+      "step": 5049
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014849999999999998,
+      "loss": 4.6815,
+      "step": 5050
+    },
+    {
+      "epoch": 8.0816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014834999999999997,
+      "loss": 3.4338,
+      "step": 5051
+    },
+    {
+      "epoch": 8.0832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001482,
+      "loss": 4.3434,
+      "step": 5052
+    },
+    {
+      "epoch": 8.0848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014805,
+      "loss": 3.0588,
+      "step": 5053
+    },
+    {
+      "epoch": 8.0864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001479,
+      "loss": 3.5971,
+      "step": 5054
+    },
+    {
+      "epoch": 8.088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014774999999999999,
+      "loss": 3.661,
+      "step": 5055
+    },
+    {
+      "epoch": 8.0896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014759999999999998,
+      "loss": 4.0386,
+      "step": 5056
+    },
+    {
+      "epoch": 8.0912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014744999999999998,
+      "loss": 3.4621,
+      "step": 5057
+    },
+    {
+      "epoch": 8.0928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014729999999999998,
+      "loss": 3.5794,
+      "step": 5058
+    },
+    {
+      "epoch": 8.0944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014714999999999997,
+      "loss": 2.784,
+      "step": 5059
+    },
+    {
+      "epoch": 8.096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000147,
+      "loss": 2.583,
+      "step": 5060
+    },
+    {
+      "epoch": 8.0976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014685,
+      "loss": 2.8943,
+      "step": 5061
+    },
+    {
+      "epoch": 8.0992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001467,
+      "loss": 2.8541,
+      "step": 5062
+    },
+    {
+      "epoch": 8.1008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014654999999999998,
+      "loss": 2.8582,
+      "step": 5063
+    },
+    {
+      "epoch": 8.1024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014639999999999998,
+      "loss": 3.1044,
+      "step": 5064
+    },
+    {
+      "epoch": 8.104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014624999999999998,
+      "loss": 3.0776,
+      "step": 5065
+    },
+    {
+      "epoch": 8.1056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014609999999999997,
+      "loss": 3.434,
+      "step": 5066
+    },
+    {
+      "epoch": 8.1072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014595,
+      "loss": 3.0661,
+      "step": 5067
+    },
+    {
+      "epoch": 8.1088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001458,
+      "loss": 3.0772,
+      "step": 5068
+    },
+    {
+      "epoch": 8.1104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014565,
+      "loss": 3.5264,
+      "step": 5069
+    },
+    {
+      "epoch": 8.112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014549999999999999,
+      "loss": 2.9278,
+      "step": 5070
+    },
+    {
+      "epoch": 8.1136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014534999999999998,
+      "loss": 2.7835,
+      "step": 5071
+    },
+    {
+      "epoch": 8.1152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014519999999999998,
+      "loss": 2.7639,
+      "step": 5072
+    },
+    {
+      "epoch": 8.1168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014504999999999997,
+      "loss": 2.7795,
+      "step": 5073
+    },
+    {
+      "epoch": 8.1184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014489999999999997,
+      "loss": 2.9886,
+      "step": 5074
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014475,
+      "loss": 2.7629,
+      "step": 5075
+    },
+    {
+      "epoch": 8.1216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001446,
+      "loss": 2.7788,
+      "step": 5076
+    },
+    {
+      "epoch": 8.1232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014445,
+      "loss": 3.0377,
+      "step": 5077
+    },
+    {
+      "epoch": 8.1248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014429999999999998,
+      "loss": 2.5952,
+      "step": 5078
+    },
+    {
+      "epoch": 8.1264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014414999999999998,
+      "loss": 2.4114,
+      "step": 5079
+    },
+    {
+      "epoch": 8.128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 2.8515,
+      "step": 5080
+    },
+    {
+      "epoch": 8.1296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014384999999999997,
+      "loss": 2.9155,
+      "step": 5081
+    },
+    {
+      "epoch": 8.1312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014369999999999997,
+      "loss": 3.2422,
+      "step": 5082
+    },
+    {
+      "epoch": 8.1328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014355,
+      "loss": 2.7588,
+      "step": 5083
+    },
+    {
+      "epoch": 8.1344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001434,
+      "loss": 2.8192,
+      "step": 5084
+    },
+    {
+      "epoch": 8.136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014324999999999999,
+      "loss": 2.9131,
+      "step": 5085
+    },
+    {
+      "epoch": 8.1376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014309999999999998,
+      "loss": 2.2079,
+      "step": 5086
+    },
+    {
+      "epoch": 8.1392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014294999999999998,
+      "loss": 3.2466,
+      "step": 5087
+    },
+    {
+      "epoch": 8.1408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014279999999999997,
+      "loss": 2.9861,
+      "step": 5088
+    },
+    {
+      "epoch": 8.1424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014264999999999997,
+      "loss": 2.6862,
+      "step": 5089
+    },
+    {
+      "epoch": 8.144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001425,
+      "loss": 2.6454,
+      "step": 5090
+    },
+    {
+      "epoch": 8.1456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014235,
+      "loss": 2.9265,
+      "step": 5091
+    },
+    {
+      "epoch": 8.1472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001422,
+      "loss": 3.5343,
+      "step": 5092
+    },
+    {
+      "epoch": 8.1488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014204999999999998,
+      "loss": 2.8883,
+      "step": 5093
+    },
+    {
+      "epoch": 8.1504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014189999999999998,
+      "loss": 3.3708,
+      "step": 5094
+    },
+    {
+      "epoch": 8.152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014174999999999998,
+      "loss": 3.6141,
+      "step": 5095
+    },
+    {
+      "epoch": 8.1536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014159999999999997,
+      "loss": 2.6007,
+      "step": 5096
+    },
+    {
+      "epoch": 8.1552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014144999999999997,
+      "loss": 2.7028,
+      "step": 5097
+    },
+    {
+      "epoch": 8.1568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001413,
+      "loss": 3.3034,
+      "step": 5098
+    },
+    {
+      "epoch": 8.1584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014115,
+      "loss": 3.0746,
+      "step": 5099
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 3.4602,
+      "step": 5100
+    },
+    {
+      "epoch": 8.1616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014084999999999998,
+      "loss": 4.0837,
+      "step": 5101
+    },
+    {
+      "epoch": 8.1632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014069999999999998,
+      "loss": 3.8424,
+      "step": 5102
+    },
+    {
+      "epoch": 8.1648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014055,
+      "loss": 5.0357,
+      "step": 5103
+    },
+    {
+      "epoch": 8.1664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001404,
+      "loss": 3.3535,
+      "step": 5104
+    },
+    {
+      "epoch": 8.168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014025,
+      "loss": 2.9491,
+      "step": 5105
+    },
+    {
+      "epoch": 8.1696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001401,
+      "loss": 3.0972,
+      "step": 5106
+    },
+    {
+      "epoch": 8.1712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013995,
+      "loss": 2.8251,
+      "step": 5107
+    },
+    {
+      "epoch": 8.1728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013979999999999998,
+      "loss": 2.6972,
+      "step": 5108
+    },
+    {
+      "epoch": 8.1744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013965,
+      "loss": 2.8575,
+      "step": 5109
+    },
+    {
+      "epoch": 8.176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001395,
+      "loss": 2.9295,
+      "step": 5110
+    },
+    {
+      "epoch": 8.1776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013935,
+      "loss": 2.8582,
+      "step": 5111
+    },
+    {
+      "epoch": 8.1792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001392,
+      "loss": 3.0131,
+      "step": 5112
+    },
+    {
+      "epoch": 8.1808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013905,
+      "loss": 2.8202,
+      "step": 5113
+    },
+    {
+      "epoch": 8.1824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001389,
+      "loss": 2.2841,
+      "step": 5114
+    },
+    {
+      "epoch": 8.184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013874999999999998,
+      "loss": 2.8865,
+      "step": 5115
+    },
+    {
+      "epoch": 8.1856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001386,
+      "loss": 3.2878,
+      "step": 5116
+    },
+    {
+      "epoch": 8.1872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013845,
+      "loss": 2.5579,
+      "step": 5117
+    },
+    {
+      "epoch": 8.1888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001383,
+      "loss": 3.4639,
+      "step": 5118
+    },
+    {
+      "epoch": 8.1904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013815,
+      "loss": 2.1206,
+      "step": 5119
+    },
+    {
+      "epoch": 8.192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000138,
+      "loss": 3.2176,
+      "step": 5120
+    },
+    {
+      "epoch": 8.1936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013785,
+      "loss": 3.531,
+      "step": 5121
+    },
+    {
+      "epoch": 8.1952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013769999999999999,
+      "loss": 2.5251,
+      "step": 5122
+    },
+    {
+      "epoch": 8.1968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013754999999999998,
+      "loss": 2.7042,
+      "step": 5123
+    },
+    {
+      "epoch": 8.1984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001374,
+      "loss": 2.6401,
+      "step": 5124
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013725,
+      "loss": 2.8847,
+      "step": 5125
+    },
+    {
+      "epoch": 8.2016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001371,
+      "loss": 2.831,
+      "step": 5126
+    },
+    {
+      "epoch": 8.2032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013695,
+      "loss": 3.003,
+      "step": 5127
+    },
+    {
+      "epoch": 8.2048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001368,
+      "loss": 3.5552,
+      "step": 5128
+    },
+    {
+      "epoch": 8.2064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013665,
+      "loss": 3.0249,
+      "step": 5129
+    },
+    {
+      "epoch": 8.208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013649999999999998,
+      "loss": 2.6,
+      "step": 5130
+    },
+    {
+      "epoch": 8.2096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013634999999999998,
+      "loss": 3.1531,
+      "step": 5131
+    },
+    {
+      "epoch": 8.2112,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001362,
+      "loss": 3.5822,
+      "step": 5132
+    },
+    {
+      "epoch": 8.2128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013605,
+      "loss": 2.864,
+      "step": 5133
+    },
+    {
+      "epoch": 8.2144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001359,
+      "loss": 3.009,
+      "step": 5134
+    },
+    {
+      "epoch": 8.216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013575,
+      "loss": 2.92,
+      "step": 5135
+    },
+    {
+      "epoch": 8.2176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001356,
+      "loss": 2.5759,
+      "step": 5136
+    },
+    {
+      "epoch": 8.2192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013544999999999999,
+      "loss": 3.3534,
+      "step": 5137
+    },
+    {
+      "epoch": 8.2208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013529999999999998,
+      "loss": 2.4112,
+      "step": 5138
+    },
+    {
+      "epoch": 8.2224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013514999999999998,
+      "loss": 3.7695,
+      "step": 5139
+    },
+    {
+      "epoch": 8.224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000135,
+      "loss": 3.4061,
+      "step": 5140
+    },
+    {
+      "epoch": 8.2256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013485,
+      "loss": 2.9928,
+      "step": 5141
+    },
+    {
+      "epoch": 8.2272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001347,
+      "loss": 2.9372,
+      "step": 5142
+    },
+    {
+      "epoch": 8.2288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013455,
+      "loss": 2.407,
+      "step": 5143
+    },
+    {
+      "epoch": 8.2304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001344,
+      "loss": 2.7185,
+      "step": 5144
+    },
+    {
+      "epoch": 8.232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013424999999999998,
+      "loss": 3.0936,
+      "step": 5145
+    },
+    {
+      "epoch": 8.2336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013409999999999998,
+      "loss": 3.018,
+      "step": 5146
+    },
+    {
+      "epoch": 8.2352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013395,
+      "loss": 2.9415,
+      "step": 5147
+    },
+    {
+      "epoch": 8.2368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001338,
+      "loss": 3.6344,
+      "step": 5148
+    },
+    {
+      "epoch": 8.2384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013365,
+      "loss": 3.3231,
+      "step": 5149
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001335,
+      "loss": 4.426,
+      "step": 5150
+    },
+    {
+      "epoch": 8.2416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013335,
+      "loss": 4.2659,
+      "step": 5151
+    },
+    {
+      "epoch": 8.2432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013319999999999999,
+      "loss": 4.1716,
+      "step": 5152
+    },
+    {
+      "epoch": 8.2448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013304999999999998,
+      "loss": 3.4874,
+      "step": 5153
+    },
+    {
+      "epoch": 8.2464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013289999999999998,
+      "loss": 3.6386,
+      "step": 5154
+    },
+    {
+      "epoch": 8.248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013275,
+      "loss": 3.1903,
+      "step": 5155
+    },
+    {
+      "epoch": 8.2496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001326,
+      "loss": 2.9343,
+      "step": 5156
+    },
+    {
+      "epoch": 8.2512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013245,
+      "loss": 3.3521,
+      "step": 5157
+    },
+    {
+      "epoch": 8.2528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001323,
+      "loss": 3.9208,
+      "step": 5158
+    },
+    {
+      "epoch": 8.2544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013215,
+      "loss": 3.5965,
+      "step": 5159
+    },
+    {
+      "epoch": 8.256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 3.196,
+      "step": 5160
+    },
+    {
+      "epoch": 8.2576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013184999999999998,
+      "loss": 2.6157,
+      "step": 5161
+    },
+    {
+      "epoch": 8.2592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013169999999999998,
+      "loss": 3.261,
+      "step": 5162
+    },
+    {
+      "epoch": 8.2608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013155,
+      "loss": 2.9177,
+      "step": 5163
+    },
+    {
+      "epoch": 8.2624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001314,
+      "loss": 2.8547,
+      "step": 5164
+    },
+    {
+      "epoch": 8.264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013125,
+      "loss": 2.9475,
+      "step": 5165
+    },
+    {
+      "epoch": 8.2656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001311,
+      "loss": 2.9541,
+      "step": 5166
+    },
+    {
+      "epoch": 8.2672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013094999999999998,
+      "loss": 3.7245,
+      "step": 5167
+    },
+    {
+      "epoch": 8.2688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013079999999999998,
+      "loss": 2.5799,
+      "step": 5168
+    },
+    {
+      "epoch": 8.2704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013064999999999998,
+      "loss": 3.0271,
+      "step": 5169
+    },
+    {
+      "epoch": 8.272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001305,
+      "loss": 3.2876,
+      "step": 5170
+    },
+    {
+      "epoch": 8.2736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013035,
+      "loss": 3.2062,
+      "step": 5171
+    },
+    {
+      "epoch": 8.2752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001302,
+      "loss": 2.895,
+      "step": 5172
+    },
+    {
+      "epoch": 8.2768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013005,
+      "loss": 2.8953,
+      "step": 5173
+    },
+    {
+      "epoch": 8.2784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012989999999999999,
+      "loss": 2.7017,
+      "step": 5174
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012974999999999998,
+      "loss": 2.5641,
+      "step": 5175
+    },
+    {
+      "epoch": 8.2816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012959999999999998,
+      "loss": 2.3601,
+      "step": 5176
+    },
+    {
+      "epoch": 8.2832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012944999999999998,
+      "loss": 2.7067,
+      "step": 5177
+    },
+    {
+      "epoch": 8.2848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001293,
+      "loss": 3.2428,
+      "step": 5178
+    },
+    {
+      "epoch": 8.2864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012915,
+      "loss": 2.6421,
+      "step": 5179
+    },
+    {
+      "epoch": 8.288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000129,
+      "loss": 2.8029,
+      "step": 5180
+    },
+    {
+      "epoch": 8.2896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012885,
+      "loss": 2.2514,
+      "step": 5181
+    },
+    {
+      "epoch": 8.2912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012869999999999998,
+      "loss": 2.3215,
+      "step": 5182
+    },
+    {
+      "epoch": 8.2928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012854999999999998,
+      "loss": 2.6813,
+      "step": 5183
+    },
+    {
+      "epoch": 8.2944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012839999999999998,
+      "loss": 3.8293,
+      "step": 5184
+    },
+    {
+      "epoch": 8.296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012824999999999997,
+      "loss": 4.1989,
+      "step": 5185
+    },
+    {
+      "epoch": 8.2976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001281,
+      "loss": 2.7688,
+      "step": 5186
+    },
+    {
+      "epoch": 8.2992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012795,
+      "loss": 3.1238,
+      "step": 5187
+    },
+    {
+      "epoch": 8.3008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001278,
+      "loss": 3.05,
+      "step": 5188
+    },
+    {
+      "epoch": 8.3024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012764999999999999,
+      "loss": 2.8148,
+      "step": 5189
+    },
+    {
+      "epoch": 8.304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012749999999999998,
+      "loss": 2.9786,
+      "step": 5190
+    },
+    {
+      "epoch": 8.3056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012734999999999998,
+      "loss": 3.1833,
+      "step": 5191
+    },
+    {
+      "epoch": 8.3072,
+      "grad_norm": NaN,
+      "learning_rate": 0.00012734999999999998,
+      "loss": 3.6253,
+      "step": 5192
+    },
+    {
+      "epoch": 8.3088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012719999999999997,
+      "loss": 2.4153,
+      "step": 5193
+    },
+    {
+      "epoch": 8.3104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012705,
+      "loss": 3.1281,
+      "step": 5194
+    },
+    {
+      "epoch": 8.312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001269,
+      "loss": 2.9595,
+      "step": 5195
+    },
+    {
+      "epoch": 8.3136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012675,
+      "loss": 3.495,
+      "step": 5196
+    },
+    {
+      "epoch": 8.3152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001266,
+      "loss": 3.3463,
+      "step": 5197
+    },
+    {
+      "epoch": 8.3168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012644999999999998,
+      "loss": 3.6449,
+      "step": 5198
+    },
+    {
+      "epoch": 8.3184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012629999999999998,
+      "loss": 2.7316,
+      "step": 5199
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": NaN,
+      "learning_rate": 0.00012629999999999998,
+      "loss": 3.3002,
+      "step": 5200
+    },
+    {
+      "epoch": 8.3216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012614999999999998,
+      "loss": 4.1094,
+      "step": 5201
+    },
+    {
+      "epoch": 8.3232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 3.6867,
+      "step": 5202
+    },
+    {
+      "epoch": 8.3248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012585,
+      "loss": 3.3427,
+      "step": 5203
+    },
+    {
+      "epoch": 8.3264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001257,
+      "loss": 4.1188,
+      "step": 5204
+    },
+    {
+      "epoch": 8.328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012555,
+      "loss": 3.5926,
+      "step": 5205
+    },
+    {
+      "epoch": 8.3296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012539999999999999,
+      "loss": 4.8951,
+      "step": 5206
+    },
+    {
+      "epoch": 8.3312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012524999999999998,
+      "loss": 3.4692,
+      "step": 5207
+    },
+    {
+      "epoch": 8.3328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012509999999999998,
+      "loss": 3.5879,
+      "step": 5208
+    },
+    {
+      "epoch": 8.3344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012494999999999997,
+      "loss": 3.5962,
+      "step": 5209
+    },
+    {
+      "epoch": 8.336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012479999999999997,
+      "loss": 3.2599,
+      "step": 5210
+    },
+    {
+      "epoch": 8.3376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012465,
+      "loss": 3.4111,
+      "step": 5211
+    },
+    {
+      "epoch": 8.3392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001245,
+      "loss": 2.5443,
+      "step": 5212
+    },
+    {
+      "epoch": 8.3408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012435,
+      "loss": 2.949,
+      "step": 5213
+    },
+    {
+      "epoch": 8.3424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012419999999999998,
+      "loss": 2.4854,
+      "step": 5214
+    },
+    {
+      "epoch": 8.344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012404999999999998,
+      "loss": 2.4217,
+      "step": 5215
+    },
+    {
+      "epoch": 8.3456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012389999999999998,
+      "loss": 2.8774,
+      "step": 5216
+    },
+    {
+      "epoch": 8.3472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012374999999999997,
+      "loss": 2.7447,
+      "step": 5217
+    },
+    {
+      "epoch": 8.3488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001236,
+      "loss": 2.4299,
+      "step": 5218
+    },
+    {
+      "epoch": 8.3504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012345,
+      "loss": 3.1645,
+      "step": 5219
+    },
+    {
+      "epoch": 8.352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001233,
+      "loss": 3.019,
+      "step": 5220
+    },
+    {
+      "epoch": 8.3536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012314999999999998,
+      "loss": 3.0103,
+      "step": 5221
+    },
+    {
+      "epoch": 8.3552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 2.2844,
+      "step": 5222
+    },
+    {
+      "epoch": 8.3568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012284999999999998,
+      "loss": 3.2206,
+      "step": 5223
+    },
+    {
+      "epoch": 8.3584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012269999999999997,
+      "loss": 2.6892,
+      "step": 5224
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012254999999999997,
+      "loss": 2.8515,
+      "step": 5225
+    },
+    {
+      "epoch": 8.3616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001224,
+      "loss": 3.5186,
+      "step": 5226
+    },
+    {
+      "epoch": 8.3632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012225,
+      "loss": 2.4005,
+      "step": 5227
+    },
+    {
+      "epoch": 8.3648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012209999999999999,
+      "loss": 3.0731,
+      "step": 5228
+    },
+    {
+      "epoch": 8.3664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012194999999999998,
+      "loss": 2.9212,
+      "step": 5229
+    },
+    {
+      "epoch": 8.368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012179999999999999,
+      "loss": 3.3993,
+      "step": 5230
+    },
+    {
+      "epoch": 8.3696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012165,
+      "loss": 2.7038,
+      "step": 5231
+    },
+    {
+      "epoch": 8.3712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001215,
+      "loss": 2.5096,
+      "step": 5232
+    },
+    {
+      "epoch": 8.3728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012135,
+      "loss": 2.288,
+      "step": 5233
+    },
+    {
+      "epoch": 8.3744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012119999999999999,
+      "loss": 2.904,
+      "step": 5234
+    },
+    {
+      "epoch": 8.376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012105,
+      "loss": 2.927,
+      "step": 5235
+    },
+    {
+      "epoch": 8.3776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001209,
+      "loss": 3.1311,
+      "step": 5236
+    },
+    {
+      "epoch": 8.3792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012075,
+      "loss": 2.6616,
+      "step": 5237
+    },
+    {
+      "epoch": 8.3808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012059999999999999,
+      "loss": 3.3794,
+      "step": 5238
+    },
+    {
+      "epoch": 8.3824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012045,
+      "loss": 2.8368,
+      "step": 5239
+    },
+    {
+      "epoch": 8.384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001203,
+      "loss": 3.2011,
+      "step": 5240
+    },
+    {
+      "epoch": 8.3856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012014999999999999,
+      "loss": 2.7275,
+      "step": 5241
+    },
+    {
+      "epoch": 8.3872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 2.8589,
+      "step": 5242
+    },
+    {
+      "epoch": 8.3888,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011985,
+      "loss": 3.5764,
+      "step": 5243
+    },
+    {
+      "epoch": 8.3904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001197,
+      "loss": 2.4109,
+      "step": 5244
+    },
+    {
+      "epoch": 8.392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011954999999999999,
+      "loss": 2.4054,
+      "step": 5245
+    },
+    {
+      "epoch": 8.3936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001194,
+      "loss": 2.9462,
+      "step": 5246
+    },
+    {
+      "epoch": 8.395199999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011925,
+      "loss": 3.132,
+      "step": 5247
+    },
+    {
+      "epoch": 8.3968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001191,
+      "loss": 3.2663,
+      "step": 5248
+    },
+    {
+      "epoch": 8.3984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011894999999999999,
+      "loss": 3.5332,
+      "step": 5249
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": NaN,
+      "learning_rate": 0.00011894999999999999,
+      "loss": 2.3482,
+      "step": 5250
+    },
+    {
+      "epoch": 8.4016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001188,
+      "loss": 4.2295,
+      "step": 5251
+    },
+    {
+      "epoch": 8.4032,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011865,
+      "loss": 3.4233,
+      "step": 5252
+    },
+    {
+      "epoch": 8.4048,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001185,
+      "loss": 3.2728,
+      "step": 5253
+    },
+    {
+      "epoch": 8.4064,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011834999999999999,
+      "loss": 3.4883,
+      "step": 5254
+    },
+    {
+      "epoch": 8.408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001182,
+      "loss": 3.0078,
+      "step": 5255
+    },
+    {
+      "epoch": 8.4096,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011805,
+      "loss": 3.2885,
+      "step": 5256
+    },
+    {
+      "epoch": 8.411200000000001,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011789999999999999,
+      "loss": 2.7993,
+      "step": 5257
+    },
+    {
+      "epoch": 8.4128,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011774999999999999,
+      "loss": 3.1993,
+      "step": 5258
+    },
+    {
+      "epoch": 8.4144,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001176,
+      "loss": 3.2918,
+      "step": 5259
+    },
+    {
+      "epoch": 8.416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011745,
+      "loss": 3.0677,
+      "step": 5260
+    },
+    {
+      "epoch": 8.4176,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011729999999999999,
+      "loss": 2.7578,
+      "step": 5261
+    },
+    {
+      "epoch": 8.4192,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011714999999999999,
+      "loss": 2.841,
+      "step": 5262
+    },
+    {
+      "epoch": 8.4208,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000117,
+      "loss": 3.4409,
+      "step": 5263
+    },
+    {
+      "epoch": 8.4224,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011685,
+      "loss": 2.457,
+      "step": 5264
+    },
+    {
+      "epoch": 8.424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011669999999999999,
+      "loss": 2.8304,
+      "step": 5265
+    },
+    {
+      "epoch": 8.4256,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011654999999999999,
+      "loss": 2.9242,
+      "step": 5266
+    },
+    {
+      "epoch": 8.4272,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001164,
+      "loss": 3.1701,
+      "step": 5267
+    },
+    {
+      "epoch": 8.4288,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011624999999999999,
+      "loss": 2.2805,
+      "step": 5268
+    },
+    {
+      "epoch": 8.4304,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011609999999999999,
+      "loss": 3.0014,
+      "step": 5269
+    },
+    {
+      "epoch": 8.432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011595,
+      "loss": 3.1696,
+      "step": 5270
+    },
+    {
+      "epoch": 8.4336,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001158,
+      "loss": 2.6706,
+      "step": 5271
+    },
+    {
+      "epoch": 8.4352,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011564999999999999,
+      "loss": 2.7588,
+      "step": 5272
+    },
+    {
+      "epoch": 8.4368,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011549999999999999,
+      "loss": 2.5778,
+      "step": 5273
+    },
+    {
+      "epoch": 8.4384,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011535,
+      "loss": 2.6264,
+      "step": 5274
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001152,
+      "loss": 3.1378,
+      "step": 5275
+    },
+    {
+      "epoch": 8.4416,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011504999999999999,
+      "loss": 2.9581,
+      "step": 5276
+    },
+    {
+      "epoch": 8.4432,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011489999999999999,
+      "loss": 2.5971,
+      "step": 5277
+    },
+    {
+      "epoch": 8.4448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011475,
+      "loss": 2.7245,
+      "step": 5278
+    },
+    {
+      "epoch": 8.4464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001146,
+      "loss": 3.5557,
+      "step": 5279
+    },
+    {
+      "epoch": 8.448,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011444999999999999,
+      "loss": 2.7126,
+      "step": 5280
+    },
+    {
+      "epoch": 8.4496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011429999999999999,
+      "loss": 2.6388,
+      "step": 5281
+    },
+    {
+      "epoch": 8.4512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011415,
+      "loss": 3.2968,
+      "step": 5282
+    },
+    {
+      "epoch": 8.4528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 2.8465,
+      "step": 5283
+    },
+    {
+      "epoch": 8.4544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011384999999999999,
+      "loss": 3.0418,
+      "step": 5284
+    },
+    {
+      "epoch": 8.456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011369999999999999,
+      "loss": 4.1577,
+      "step": 5285
+    },
+    {
+      "epoch": 8.4576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011355,
+      "loss": 2.5981,
+      "step": 5286
+    },
+    {
+      "epoch": 8.4592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011339999999999999,
+      "loss": 2.8671,
+      "step": 5287
+    },
+    {
+      "epoch": 8.4608,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011324999999999999,
+      "loss": 3.0664,
+      "step": 5288
+    },
+    {
+      "epoch": 8.4624,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011309999999999998,
+      "loss": 2.6248,
+      "step": 5289
+    },
+    {
+      "epoch": 8.464,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011295,
+      "loss": 3.144,
+      "step": 5290
+    },
+    {
+      "epoch": 8.4656,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011279999999999999,
+      "loss": 2.5095,
+      "step": 5291
+    },
+    {
+      "epoch": 8.4672,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011264999999999999,
+      "loss": 2.9168,
+      "step": 5292
+    },
+    {
+      "epoch": 8.4688,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001125,
+      "loss": 3.5242,
+      "step": 5293
+    },
+    {
+      "epoch": 8.4704,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011235,
+      "loss": 3.1218,
+      "step": 5294
+    },
+    {
+      "epoch": 8.472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011219999999999999,
+      "loss": 3.3528,
+      "step": 5295
+    },
+    {
+      "epoch": 8.4736,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011204999999999999,
+      "loss": 3.208,
+      "step": 5296
+    },
+    {
+      "epoch": 8.4752,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001119,
+      "loss": 3.2412,
+      "step": 5297
+    },
+    {
+      "epoch": 8.4768,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011174999999999999,
+      "loss": 3.0845,
+      "step": 5298
+    },
+    {
+      "epoch": 8.4784,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011159999999999999,
+      "loss": 2.8777,
+      "step": 5299
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011144999999999998,
+      "loss": 3.3584,
+      "step": 5300
+    },
+    {
+      "epoch": 8.4816,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001113,
+      "loss": 5.1918,
+      "step": 5301
+    },
+    {
+      "epoch": 8.4832,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011114999999999999,
+      "loss": 3.8634,
+      "step": 5302
+    },
+    {
+      "epoch": 8.4848,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 3.2812,
+      "step": 5303
+    },
+    {
+      "epoch": 8.4864,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011084999999999998,
+      "loss": 3.6451,
+      "step": 5304
+    },
+    {
+      "epoch": 8.488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001107,
+      "loss": 4.7517,
+      "step": 5305
+    },
+    {
+      "epoch": 8.4896,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011054999999999999,
+      "loss": 4.3051,
+      "step": 5306
+    },
+    {
+      "epoch": 8.4912,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011039999999999999,
+      "loss": 3.1874,
+      "step": 5307
+    },
+    {
+      "epoch": 8.4928,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011024999999999998,
+      "loss": 3.5162,
+      "step": 5308
+    },
+    {
+      "epoch": 8.4944,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011009999999999999,
+      "loss": 3.6533,
+      "step": 5309
+    },
+    {
+      "epoch": 8.496,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010994999999999999,
+      "loss": 3.2079,
+      "step": 5310
+    },
+    {
+      "epoch": 8.4976,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010979999999999999,
+      "loss": 3.0578,
+      "step": 5311
+    },
+    {
+      "epoch": 8.4992,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010964999999999998,
+      "loss": 2.8665,
+      "step": 5312
+    },
+    {
+      "epoch": 8.5008,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010949999999999999,
+      "loss": 3.2601,
+      "step": 5313
+    },
+    {
+      "epoch": 8.5024,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010934999999999999,
+      "loss": 3.4261,
+      "step": 5314
+    },
+    {
+      "epoch": 8.504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010919999999999998,
+      "loss": 2.7026,
+      "step": 5315
+    },
+    {
+      "epoch": 8.5056,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010904999999999998,
+      "loss": 2.4031,
+      "step": 5316
+    },
+    {
+      "epoch": 8.5072,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010889999999999999,
+      "loss": 3.1147,
+      "step": 5317
+    },
+    {
+      "epoch": 8.5088,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010874999999999999,
+      "loss": 3.1283,
+      "step": 5318
+    },
+    {
+      "epoch": 8.5104,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010859999999999998,
+      "loss": 2.8211,
+      "step": 5319
+    },
+    {
+      "epoch": 8.512,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010845,
+      "loss": 3.076,
+      "step": 5320
+    },
+    {
+      "epoch": 8.5136,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010829999999999999,
+      "loss": 3.2892,
+      "step": 5321
+    },
+    {
+      "epoch": 8.5152,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010814999999999999,
+      "loss": 3.1918,
+      "step": 5322
+    },
+    {
+      "epoch": 8.5168,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 2.7595,
+      "step": 5323
+    },
+    {
+      "epoch": 8.5184,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010784999999999999,
+      "loss": 2.7904,
+      "step": 5324
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010769999999999999,
+      "loss": 3.2722,
+      "step": 5325
+    },
+    {
+      "epoch": 8.5216,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010754999999999999,
+      "loss": 2.5779,
+      "step": 5326
+    },
+    {
+      "epoch": 8.5232,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010739999999999998,
+      "loss": 2.7775,
+      "step": 5327
+    },
+    {
+      "epoch": 8.5248,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010724999999999999,
+      "loss": 3.1146,
+      "step": 5328
+    },
+    {
+      "epoch": 8.5264,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010709999999999999,
+      "loss": 2.633,
+      "step": 5329
+    },
+    {
+      "epoch": 8.528,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010694999999999998,
+      "loss": 2.5443,
+      "step": 5330
+    },
+    {
+      "epoch": 8.5296,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010679999999999998,
+      "loss": 3.069,
+      "step": 5331
+    },
+    {
+      "epoch": 8.5312,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010664999999999999,
+      "loss": 3.0671,
+      "step": 5332
+    },
+    {
+      "epoch": 8.5328,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010649999999999999,
+      "loss": 2.842,
+      "step": 5333
+    },
+    {
+      "epoch": 8.5344,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010634999999999998,
+      "loss": 3.2067,
+      "step": 5334
+    },
+    {
+      "epoch": 8.536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010619999999999998,
+      "loss": 2.835,
+      "step": 5335
+    },
+    {
+      "epoch": 8.5376,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010604999999999999,
+      "loss": 2.7643,
+      "step": 5336
+    },
+    {
+      "epoch": 8.5392,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010589999999999999,
+      "loss": 2.9063,
+      "step": 5337
+    },
+    {
+      "epoch": 8.5408,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010574999999999998,
+      "loss": 3.4485,
+      "step": 5338
+    },
+    {
+      "epoch": 8.5424,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010559999999999998,
+      "loss": 2.6321,
+      "step": 5339
+    },
+    {
+      "epoch": 8.544,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010544999999999999,
+      "loss": 2.9262,
+      "step": 5340
+    },
+    {
+      "epoch": 8.5456,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010529999999999998,
+      "loss": 3.3372,
+      "step": 5341
+    },
+    {
+      "epoch": 8.5472,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010514999999999998,
+      "loss": 3.9157,
+      "step": 5342
+    },
+    {
+      "epoch": 8.5488,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 2.9866,
+      "step": 5343
+    },
+    {
+      "epoch": 8.5504,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010484999999999999,
+      "loss": 1.9647,
+      "step": 5344
+    },
+    {
+      "epoch": 8.552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010469999999999998,
+      "loss": 2.4543,
+      "step": 5345
+    },
+    {
+      "epoch": 8.5536,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010454999999999998,
+      "loss": 3.2684,
+      "step": 5346
+    },
+    {
+      "epoch": 8.5552,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010439999999999999,
+      "loss": 2.7297,
+      "step": 5347
+    },
+    {
+      "epoch": 8.556799999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010424999999999999,
+      "loss": 4.141,
+      "step": 5348
+    },
+    {
+      "epoch": 8.5584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010409999999999998,
+      "loss": 2.9914,
+      "step": 5349
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010394999999999998,
+      "loss": 4.2832,
+      "step": 5350
+    },
+    {
+      "epoch": 8.5616,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010379999999999999,
+      "loss": 4.2738,
+      "step": 5351
+    },
+    {
+      "epoch": 8.5632,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010364999999999999,
+      "loss": 3.7024,
+      "step": 5352
+    },
+    {
+      "epoch": 8.5648,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010349999999999998,
+      "loss": 4.0658,
+      "step": 5353
+    },
+    {
+      "epoch": 8.5664,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010334999999999998,
+      "loss": 4.0144,
+      "step": 5354
+    },
+    {
+      "epoch": 8.568,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010319999999999999,
+      "loss": 3.2457,
+      "step": 5355
+    },
+    {
+      "epoch": 8.5696,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010305,
+      "loss": 3.4408,
+      "step": 5356
+    },
+    {
+      "epoch": 8.5712,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001029,
+      "loss": 3.0264,
+      "step": 5357
+    },
+    {
+      "epoch": 8.5728,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010275,
+      "loss": 3.8298,
+      "step": 5358
+    },
+    {
+      "epoch": 8.5744,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001026,
+      "loss": 3.779,
+      "step": 5359
+    },
+    {
+      "epoch": 8.576,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010245,
+      "loss": 3.0959,
+      "step": 5360
+    },
+    {
+      "epoch": 8.5776,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010229999999999999,
+      "loss": 3.6852,
+      "step": 5361
+    },
+    {
+      "epoch": 8.5792,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010215,
+      "loss": 3.4055,
+      "step": 5362
+    },
+    {
+      "epoch": 8.5808,
+      "grad_norm": 0.0,
+      "learning_rate": 0.000102,
+      "loss": 3.4038,
+      "step": 5363
+    },
+    {
+      "epoch": 8.5824,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010185,
+      "loss": 3.2311,
+      "step": 5364
+    },
+    {
+      "epoch": 8.584,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010169999999999999,
+      "loss": 2.9334,
+      "step": 5365
+    },
+    {
+      "epoch": 8.5856,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010155,
+      "loss": 2.3327,
+      "step": 5366
+    },
+    {
+      "epoch": 8.5872,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001014,
+      "loss": 3.4392,
+      "step": 5367
+    },
+    {
+      "epoch": 8.588799999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010125,
+      "loss": 2.688,
+      "step": 5368
+    },
+    {
+      "epoch": 8.5904,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001011,
+      "loss": 3.0311,
+      "step": 5369
+    },
+    {
+      "epoch": 8.592,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010095,
+      "loss": 3.853,
+      "step": 5370
+    },
+    {
+      "epoch": 8.5936,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001008,
+      "loss": 3.2355,
+      "step": 5371
+    },
+    {
+      "epoch": 8.5952,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010065,
+      "loss": 2.6658,
+      "step": 5372
+    },
+    {
+      "epoch": 8.5968,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001005,
+      "loss": 2.8656,
+      "step": 5373
+    },
+    {
+      "epoch": 8.5984,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010035,
+      "loss": 2.7324,
+      "step": 5374
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001002,
+      "loss": 2.9669,
+      "step": 5375
+    },
+    {
+      "epoch": 8.6016,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010004999999999999,
+      "loss": 2.1973,
+      "step": 5376
+    },
+    {
+      "epoch": 8.6032,
+      "grad_norm": 0.0,
+      "learning_rate": 9.99e-05,
+      "loss": 2.7826,
+      "step": 5377
+    },
+    {
+      "epoch": 8.604800000000001,
+      "grad_norm": 0.0,
+      "learning_rate": 9.975e-05,
+      "loss": 2.4138,
+      "step": 5378
+    },
+    {
+      "epoch": 8.6064,
+      "grad_norm": 0.0,
+      "learning_rate": 9.96e-05,
+      "loss": 2.999,
+      "step": 5379
+    },
+    {
+      "epoch": 8.608,
+      "grad_norm": 0.0,
+      "learning_rate": 9.944999999999999e-05,
+      "loss": 3.0998,
+      "step": 5380
+    },
+    {
+      "epoch": 8.6096,
+      "grad_norm": 0.0,
+      "learning_rate": 9.93e-05,
+      "loss": 2.8706,
+      "step": 5381
+    },
+    {
+      "epoch": 8.6112,
+      "grad_norm": 0.0,
+      "learning_rate": 9.915e-05,
+      "loss": 2.7169,
+      "step": 5382
+    },
+    {
+      "epoch": 8.6128,
+      "grad_norm": 0.0,
+      "learning_rate": 9.9e-05,
+      "loss": 3.1331,
+      "step": 5383
+    },
+    {
+      "epoch": 8.6144,
+      "grad_norm": 0.0,
+      "learning_rate": 9.884999999999999e-05,
+      "loss": 2.4797,
+      "step": 5384
+    },
+    {
+      "epoch": 8.616,
+      "grad_norm": 0.0,
+      "learning_rate": 9.87e-05,
+      "loss": 2.5391,
+      "step": 5385
+    },
+    {
+      "epoch": 8.6176,
+      "grad_norm": 0.0,
+      "learning_rate": 9.855e-05,
+      "loss": 2.4524,
+      "step": 5386
+    },
+    {
+      "epoch": 8.6192,
+      "grad_norm": 0.0,
+      "learning_rate": 9.839999999999999e-05,
+      "loss": 2.9776,
+      "step": 5387
+    },
+    {
+      "epoch": 8.6208,
+      "grad_norm": 0.0,
+      "learning_rate": 9.824999999999999e-05,
+      "loss": 3.2171,
+      "step": 5388
+    },
+    {
+      "epoch": 8.6224,
+      "grad_norm": 0.0,
+      "learning_rate": 9.81e-05,
+      "loss": 3.4589,
+      "step": 5389
+    },
+    {
+      "epoch": 8.624,
+      "grad_norm": 0.0,
+      "learning_rate": 9.795e-05,
+      "loss": 2.4906,
+      "step": 5390
+    },
+    {
+      "epoch": 8.6256,
+      "grad_norm": 0.0,
+      "learning_rate": 9.779999999999999e-05,
+      "loss": 3.3645,
+      "step": 5391
+    },
+    {
+      "epoch": 8.6272,
+      "grad_norm": 0.0,
+      "learning_rate": 9.764999999999999e-05,
+      "loss": 2.7121,
+      "step": 5392
+    },
+    {
+      "epoch": 8.6288,
+      "grad_norm": 0.0,
+      "learning_rate": 9.75e-05,
+      "loss": 3.3566,
+      "step": 5393
+    },
+    {
+      "epoch": 8.6304,
+      "grad_norm": 0.0,
+      "learning_rate": 9.735e-05,
+      "loss": 3.0868,
+      "step": 5394
+    },
+    {
+      "epoch": 8.632,
+      "grad_norm": 0.0,
+      "learning_rate": 9.719999999999999e-05,
+      "loss": 2.3634,
+      "step": 5395
+    },
+    {
+      "epoch": 8.6336,
+      "grad_norm": 0.0,
+      "learning_rate": 9.705e-05,
+      "loss": 2.4537,
+      "step": 5396
+    },
+    {
+      "epoch": 8.6352,
+      "grad_norm": 0.0,
+      "learning_rate": 9.69e-05,
+      "loss": 2.7268,
+      "step": 5397
+    },
+    {
+      "epoch": 8.636800000000001,
+      "grad_norm": 0.0,
+      "learning_rate": 9.675e-05,
+      "loss": 2.4014,
+      "step": 5398
+    },
+    {
+      "epoch": 8.6384,
+      "grad_norm": 0.0,
+      "learning_rate": 9.659999999999999e-05,
+      "loss": 3.4346,
+      "step": 5399
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.645e-05,
+      "loss": 3.1193,
+      "step": 5400
+    },
+    {
+      "epoch": 8.6416,
+      "grad_norm": 0.0,
+      "learning_rate": 9.63e-05,
+      "loss": 4.077,
+      "step": 5401
+    },
+    {
+      "epoch": 8.6432,
+      "grad_norm": 0.0,
+      "learning_rate": 9.614999999999999e-05,
+      "loss": 4.2156,
+      "step": 5402
+    },
+    {
+      "epoch": 8.6448,
+      "grad_norm": 0.0,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 3.1341,
+      "step": 5403
+    },
+    {
+      "epoch": 8.6464,
+      "grad_norm": 0.0,
+      "learning_rate": 9.585e-05,
+      "loss": 4.1184,
+      "step": 5404
+    },
+    {
+      "epoch": 8.648,
+      "grad_norm": 0.0,
+      "learning_rate": 9.57e-05,
+      "loss": 3.6441,
+      "step": 5405
+    },
+    {
+      "epoch": 8.6496,
+      "grad_norm": 0.0,
+      "learning_rate": 9.554999999999999e-05,
+      "loss": 3.4632,
+      "step": 5406
+    },
+    {
+      "epoch": 8.6512,
+      "grad_norm": 0.0,
+      "learning_rate": 9.539999999999999e-05,
+      "loss": 3.1794,
+      "step": 5407
+    },
+    {
+      "epoch": 8.6528,
+      "grad_norm": 0.0,
+      "learning_rate": 9.525e-05,
+      "loss": 4.5431,
+      "step": 5408
+    },
+    {
+      "epoch": 8.6544,
+      "grad_norm": 0.0,
+      "learning_rate": 9.51e-05,
+      "loss": 3.7904,
+      "step": 5409
+    },
+    {
+      "epoch": 8.656,
+      "grad_norm": 0.0,
+      "learning_rate": 9.494999999999999e-05,
+      "loss": 2.7677,
+      "step": 5410
+    },
+    {
+      "epoch": 8.6576,
+      "grad_norm": 0.0,
+      "learning_rate": 9.479999999999999e-05,
+      "loss": 2.5789,
+      "step": 5411
+    },
+    {
+      "epoch": 8.6592,
+      "grad_norm": 0.0,
+      "learning_rate": 9.465e-05,
+      "loss": 2.8011,
+      "step": 5412
+    },
+    {
+      "epoch": 8.6608,
+      "grad_norm": 0.0,
+      "learning_rate": 9.449999999999999e-05,
+      "loss": 3.3363,
+      "step": 5413
+    },
+    {
+      "epoch": 8.6624,
+      "grad_norm": 0.0,
+      "learning_rate": 9.434999999999999e-05,
+      "loss": 2.5781,
+      "step": 5414
+    },
+    {
+      "epoch": 8.664,
+      "grad_norm": 0.0,
+      "learning_rate": 9.419999999999999e-05,
+      "loss": 3.5032,
+      "step": 5415
+    },
+    {
+      "epoch": 8.6656,
+      "grad_norm": 0.0,
+      "learning_rate": 9.405e-05,
+      "loss": 2.6167,
+      "step": 5416
+    },
+    {
+      "epoch": 8.6672,
+      "grad_norm": 0.0,
+      "learning_rate": 9.389999999999999e-05,
+      "loss": 2.2998,
+      "step": 5417
+    },
+    {
+      "epoch": 8.6688,
+      "grad_norm": 0.0,
+      "learning_rate": 9.374999999999999e-05,
+      "loss": 2.5932,
+      "step": 5418
+    },
+    {
+      "epoch": 8.6704,
+      "grad_norm": 0.0,
+      "learning_rate": 9.36e-05,
+      "loss": 2.5544,
+      "step": 5419
+    },
+    {
+      "epoch": 8.672,
+      "grad_norm": 0.0,
+      "learning_rate": 9.345e-05,
+      "loss": 3.0412,
+      "step": 5420
+    },
+    {
+      "epoch": 8.6736,
+      "grad_norm": 0.0,
+      "learning_rate": 9.329999999999999e-05,
+      "loss": 3.6085,
+      "step": 5421
+    },
+    {
+      "epoch": 8.6752,
+      "grad_norm": 0.0,
+      "learning_rate": 9.314999999999999e-05,
+      "loss": 3.2467,
+      "step": 5422
+    },
+    {
+      "epoch": 8.6768,
+      "grad_norm": 0.0,
+      "learning_rate": 9.3e-05,
+      "loss": 3.0468,
+      "step": 5423
+    },
+    {
+      "epoch": 8.6784,
+      "grad_norm": 0.0,
+      "learning_rate": 9.285e-05,
+      "loss": 2.4345,
+      "step": 5424
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.269999999999999e-05,
+      "loss": 2.8213,
+      "step": 5425
+    },
+    {
+      "epoch": 8.6816,
+      "grad_norm": 0.0,
+      "learning_rate": 9.254999999999999e-05,
+      "loss": 2.8409,
+      "step": 5426
+    },
+    {
+      "epoch": 8.6832,
+      "grad_norm": 0.0,
+      "learning_rate": 9.24e-05,
+      "loss": 2.8714,
+      "step": 5427
+    },
+    {
+      "epoch": 8.6848,
+      "grad_norm": 0.0,
+      "learning_rate": 9.224999999999999e-05,
+      "loss": 2.918,
+      "step": 5428
+    },
+    {
+      "epoch": 8.6864,
+      "grad_norm": 0.0,
+      "learning_rate": 9.209999999999999e-05,
+      "loss": 3.2587,
+      "step": 5429
+    },
+    {
+      "epoch": 8.688,
+      "grad_norm": 0.0,
+      "learning_rate": 9.194999999999999e-05,
+      "loss": 3.1399,
+      "step": 5430
+    },
+    {
+      "epoch": 8.6896,
+      "grad_norm": 0.0,
+      "learning_rate": 9.18e-05,
+      "loss": 2.755,
+      "step": 5431
+    },
+    {
+      "epoch": 8.6912,
+      "grad_norm": 0.0,
+      "learning_rate": 9.164999999999999e-05,
+      "loss": 2.5296,
+      "step": 5432
+    },
+    {
+      "epoch": 8.6928,
+      "grad_norm": 0.0,
+      "learning_rate": 9.149999999999999e-05,
+      "loss": 2.8267,
+      "step": 5433
+    },
+    {
+      "epoch": 8.6944,
+      "grad_norm": 0.0,
+      "learning_rate": 9.134999999999998e-05,
+      "loss": 2.5156,
+      "step": 5434
+    },
+    {
+      "epoch": 8.696,
+      "grad_norm": 0.0,
+      "learning_rate": 9.12e-05,
+      "loss": 2.5486,
+      "step": 5435
+    },
+    {
+      "epoch": 8.6976,
+      "grad_norm": 0.0,
+      "learning_rate": 9.104999999999999e-05,
+      "loss": 2.533,
+      "step": 5436
+    },
+    {
+      "epoch": 8.6992,
+      "grad_norm": 0.0,
+      "learning_rate": 9.089999999999999e-05,
+      "loss": 2.7671,
+      "step": 5437
+    },
+    {
+      "epoch": 8.7008,
+      "grad_norm": 0.0,
+      "learning_rate": 9.074999999999998e-05,
+      "loss": 2.9547,
+      "step": 5438
+    },
+    {
+      "epoch": 8.7024,
+      "grad_norm": 0.0,
+      "learning_rate": 9.059999999999999e-05,
+      "loss": 2.9633,
+      "step": 5439
+    },
+    {
+      "epoch": 8.704,
+      "grad_norm": 0.0,
+      "learning_rate": 9.044999999999999e-05,
+      "loss": 3.3505,
+      "step": 5440
+    },
+    {
+      "epoch": 8.7056,
+      "grad_norm": 0.0,
+      "learning_rate": 9.029999999999999e-05,
+      "loss": 3.1255,
+      "step": 5441
+    },
+    {
+      "epoch": 8.7072,
+      "grad_norm": 0.0,
+      "learning_rate": 9.014999999999998e-05,
+      "loss": 3.4915,
+      "step": 5442
+    },
+    {
+      "epoch": 8.7088,
+      "grad_norm": 0.0,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 2.9845,
+      "step": 5443
+    },
+    {
+      "epoch": 8.7104,
+      "grad_norm": 0.0,
+      "learning_rate": 8.984999999999999e-05,
+      "loss": 3.0208,
+      "step": 5444
+    },
+    {
+      "epoch": 8.712,
+      "grad_norm": 0.0,
+      "learning_rate": 8.969999999999998e-05,
+      "loss": 2.9816,
+      "step": 5445
+    },
+    {
+      "epoch": 8.7136,
+      "grad_norm": 0.0,
+      "learning_rate": 8.955e-05,
+      "loss": 3.0994,
+      "step": 5446
+    },
+    {
+      "epoch": 8.7152,
+      "grad_norm": 0.0,
+      "learning_rate": 8.939999999999999e-05,
+      "loss": 2.7481,
+      "step": 5447
+    },
+    {
+      "epoch": 8.7168,
+      "grad_norm": 0.0,
+      "learning_rate": 8.924999999999999e-05,
+      "loss": 2.8729,
+      "step": 5448
+    },
+    {
+      "epoch": 8.7184,
+      "grad_norm": 0.0,
+      "learning_rate": 8.909999999999998e-05,
+      "loss": 3.4677,
+      "step": 5449
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 8.895e-05,
+      "loss": 2.262,
+      "step": 5450
+    },
+    {
+      "epoch": 8.7216,
+      "grad_norm": 0.0,
+      "learning_rate": 8.879999999999999e-05,
+      "loss": 4.1577,
+      "step": 5451
+    },
+    {
+      "epoch": 8.7232,
+      "grad_norm": 0.0,
+      "learning_rate": 8.864999999999999e-05,
+      "loss": 4.2487,
+      "step": 5452
+    },
+    {
+      "epoch": 8.7248,
+      "grad_norm": 0.0,
+      "learning_rate": 8.849999999999998e-05,
+      "loss": 3.2539,
+      "step": 5453
+    },
+    {
+      "epoch": 8.7264,
+      "grad_norm": 0.0,
+      "learning_rate": 8.834999999999999e-05,
+      "loss": 3.7699,
+      "step": 5454
+    },
+    {
+      "epoch": 8.728,
+      "grad_norm": 0.0,
+      "learning_rate": 8.819999999999999e-05,
+      "loss": 2.6604,
+      "step": 5455
+    },
+    {
+      "epoch": 8.7296,
+      "grad_norm": 0.0,
+      "learning_rate": 8.804999999999999e-05,
+      "loss": 4.3472,
+      "step": 5456
+    },
+    {
+      "epoch": 8.7312,
+      "grad_norm": 0.0,
+      "learning_rate": 8.789999999999998e-05,
+      "loss": 2.8844,
+      "step": 5457
+    },
+    {
+      "epoch": 8.7328,
+      "grad_norm": 0.0,
+      "learning_rate": 8.774999999999999e-05,
+      "loss": 3.1081,
+      "step": 5458
+    },
+    {
+      "epoch": 8.7344,
+      "grad_norm": 0.0,
+      "learning_rate": 8.759999999999999e-05,
+      "loss": 2.8376,
+      "step": 5459
+    },
+    {
+      "epoch": 8.736,
+      "grad_norm": 0.0,
+      "learning_rate": 8.744999999999998e-05,
+      "loss": 3.3077,
+      "step": 5460
+    },
+    {
+      "epoch": 8.7376,
+      "grad_norm": 0.0,
+      "learning_rate": 8.729999999999998e-05,
+      "loss": 2.8664,
+      "step": 5461
+    },
+    {
+      "epoch": 8.7392,
+      "grad_norm": 0.0,
+      "learning_rate": 8.714999999999999e-05,
+      "loss": 3.8246,
+      "step": 5462
+    },
+    {
+      "epoch": 8.7408,
+      "grad_norm": 0.0,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 2.7271,
+      "step": 5463
+    },
+    {
+      "epoch": 8.7424,
+      "grad_norm": 0.0,
+      "learning_rate": 8.684999999999998e-05,
+      "loss": 3.0011,
+      "step": 5464
+    },
+    {
+      "epoch": 8.744,
+      "grad_norm": 0.0,
+      "learning_rate": 8.669999999999998e-05,
+      "loss": 2.5552,
+      "step": 5465
+    },
+    {
+      "epoch": 8.7456,
+      "grad_norm": 0.0,
+      "learning_rate": 8.654999999999999e-05,
+      "loss": 2.7657,
+      "step": 5466
+    },
+    {
+      "epoch": 8.7472,
+      "grad_norm": 0.0,
+      "learning_rate": 8.639999999999999e-05,
+      "loss": 2.8866,
+      "step": 5467
+    },
+    {
+      "epoch": 8.7488,
+      "grad_norm": 0.0,
+      "learning_rate": 8.624999999999998e-05,
+      "loss": 3.3402,
+      "step": 5468
+    },
+    {
+      "epoch": 8.750399999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 8.609999999999999e-05,
+      "loss": 2.9473,
+      "step": 5469
+    },
+    {
+      "epoch": 8.752,
+      "grad_norm": 0.0,
+      "learning_rate": 8.594999999999999e-05,
+      "loss": 2.8834,
+      "step": 5470
+    },
+    {
+      "epoch": 8.7536,
+      "grad_norm": 0.0,
+      "learning_rate": 8.579999999999998e-05,
+      "loss": 2.3812,
+      "step": 5471
+    },
+    {
+      "epoch": 8.7552,
+      "grad_norm": 0.0,
+      "learning_rate": 8.564999999999998e-05,
+      "loss": 3.8734,
+      "step": 5472
+    },
+    {
+      "epoch": 8.7568,
+      "grad_norm": 0.0,
+      "learning_rate": 8.549999999999999e-05,
+      "loss": 2.6016,
+      "step": 5473
+    },
+    {
+      "epoch": 8.7584,
+      "grad_norm": 0.0,
+      "learning_rate": 8.534999999999999e-05,
+      "loss": 2.5534,
+      "step": 5474
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 8.519999999999998e-05,
+      "loss": 2.5744,
+      "step": 5475
+    },
+    {
+      "epoch": 8.7616,
+      "grad_norm": 0.0,
+      "learning_rate": 8.504999999999998e-05,
+      "loss": 2.7575,
+      "step": 5476
+    },
+    {
+      "epoch": 8.7632,
+      "grad_norm": 0.0,
+      "learning_rate": 8.489999999999999e-05,
+      "loss": 2.1876,
+      "step": 5477
+    },
+    {
+      "epoch": 8.7648,
+      "grad_norm": 0.0,
+      "learning_rate": 8.474999999999999e-05,
+      "loss": 2.4229,
+      "step": 5478
+    },
+    {
+      "epoch": 8.7664,
+      "grad_norm": 0.0,
+      "learning_rate": 8.459999999999998e-05,
+      "loss": 3.0801,
+      "step": 5479
+    },
+    {
+      "epoch": 8.768,
+      "grad_norm": 0.0,
+      "learning_rate": 8.444999999999998e-05,
+      "loss": 2.9251,
+      "step": 5480
+    },
+    {
+      "epoch": 8.7696,
+      "grad_norm": 0.0,
+      "learning_rate": 8.43e-05,
+      "loss": 2.48,
+      "step": 5481
+    },
+    {
+      "epoch": 8.7712,
+      "grad_norm": 0.0,
+      "learning_rate": 8.415e-05,
+      "loss": 2.5742,
+      "step": 5482
+    },
+    {
+      "epoch": 8.7728,
+      "grad_norm": 0.0,
+      "learning_rate": 8.4e-05,
+      "loss": 2.8933,
+      "step": 5483
+    },
+    {
+      "epoch": 8.7744,
+      "grad_norm": 0.0,
+      "learning_rate": 8.385e-05,
+      "loss": 2.7115,
+      "step": 5484
+    },
+    {
+      "epoch": 8.776,
+      "grad_norm": 0.0,
+      "learning_rate": 8.37e-05,
+      "loss": 2.6023,
+      "step": 5485
+    },
+    {
+      "epoch": 8.7776,
+      "grad_norm": 0.0,
+      "learning_rate": 8.355e-05,
+      "loss": 3.4378,
+      "step": 5486
+    },
+    {
+      "epoch": 8.7792,
+      "grad_norm": 0.0,
+      "learning_rate": 8.34e-05,
+      "loss": 3.2154,
+      "step": 5487
+    },
+    {
+      "epoch": 8.7808,
+      "grad_norm": 0.0,
+      "learning_rate": 8.325e-05,
+      "loss": 3.0487,
+      "step": 5488
+    },
+    {
+      "epoch": 8.782399999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 8.31e-05,
+      "loss": 2.8643,
+      "step": 5489
+    },
+    {
+      "epoch": 8.784,
+      "grad_norm": 0.0,
+      "learning_rate": 8.295e-05,
+      "loss": 2.6776,
+      "step": 5490
+    },
+    {
+      "epoch": 8.7856,
+      "grad_norm": 0.0,
+      "learning_rate": 8.28e-05,
+      "loss": 2.92,
+      "step": 5491
+    },
+    {
+      "epoch": 8.7872,
+      "grad_norm": 0.0,
+      "learning_rate": 8.265e-05,
+      "loss": 3.2343,
+      "step": 5492
+    },
+    {
+      "epoch": 8.7888,
+      "grad_norm": 0.0,
+      "learning_rate": 8.25e-05,
+      "loss": 2.5515,
+      "step": 5493
+    },
+    {
+      "epoch": 8.7904,
+      "grad_norm": 0.0,
+      "learning_rate": 8.235e-05,
+      "loss": 2.9681,
+      "step": 5494
+    },
+    {
+      "epoch": 8.792,
+      "grad_norm": 0.0,
+      "learning_rate": 8.22e-05,
+      "loss": 2.7059,
+      "step": 5495
+    },
+    {
+      "epoch": 8.7936,
+      "grad_norm": 0.0,
+      "learning_rate": 8.205e-05,
+      "loss": 2.2784,
+      "step": 5496
+    },
+    {
+      "epoch": 8.7952,
+      "grad_norm": 0.0,
+      "learning_rate": 8.19e-05,
+      "loss": 3.4781,
+      "step": 5497
+    },
+    {
+      "epoch": 8.7968,
+      "grad_norm": 0.0,
+      "learning_rate": 8.175e-05,
+      "loss": 3.1366,
+      "step": 5498
+    },
+    {
+      "epoch": 8.7984,
+      "grad_norm": 0.0,
+      "learning_rate": 8.16e-05,
+      "loss": 3.1494,
+      "step": 5499
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": NaN,
+      "learning_rate": 8.16e-05,
+      "loss": 3.1003,
+      "step": 5500
+    },
+    {
+      "epoch": 8.8016,
+      "grad_norm": 0.0,
+      "learning_rate": 8.145e-05,
+      "loss": 3.3448,
+      "step": 5501
+    },
+    {
+      "epoch": 8.8032,
+      "grad_norm": 0.0,
+      "learning_rate": 8.13e-05,
+      "loss": 4.0856,
+      "step": 5502
+    },
+    {
+      "epoch": 8.8048,
+      "grad_norm": 0.0,
+      "learning_rate": 8.115e-05,
+      "loss": 3.9545,
+      "step": 5503
+    },
+    {
+      "epoch": 8.8064,
+      "grad_norm": 0.0,
+      "learning_rate": 8.1e-05,
+      "loss": 3.7369,
+      "step": 5504
+    },
+    {
+      "epoch": 8.808,
+      "grad_norm": 0.0,
+      "learning_rate": 8.085e-05,
+      "loss": 2.977,
+      "step": 5505
+    },
+    {
+      "epoch": 8.8096,
+      "grad_norm": 0.0,
+      "learning_rate": 8.07e-05,
+      "loss": 3.2168,
+      "step": 5506
+    },
+    {
+      "epoch": 8.8112,
+      "grad_norm": 0.0,
+      "learning_rate": 8.054999999999999e-05,
+      "loss": 3.0844,
+      "step": 5507
+    },
+    {
+      "epoch": 8.8128,
+      "grad_norm": 0.0,
+      "learning_rate": 8.04e-05,
+      "loss": 2.8623,
+      "step": 5508
+    },
+    {
+      "epoch": 8.8144,
+      "grad_norm": 0.0,
+      "learning_rate": 8.025e-05,
+      "loss": 3.3578,
+      "step": 5509
+    },
+    {
+      "epoch": 8.816,
+      "grad_norm": 0.0,
+      "learning_rate": 8.01e-05,
+      "loss": 2.7042,
+      "step": 5510
+    },
+    {
+      "epoch": 8.8176,
+      "grad_norm": 0.0,
+      "learning_rate": 7.994999999999999e-05,
+      "loss": 3.3687,
+      "step": 5511
+    },
+    {
+      "epoch": 8.8192,
+      "grad_norm": 0.0,
+      "learning_rate": 7.98e-05,
+      "loss": 3.6682,
+      "step": 5512
+    },
+    {
+      "epoch": 8.8208,
+      "grad_norm": 0.0,
+      "learning_rate": 7.965e-05,
+      "loss": 2.6471,
+      "step": 5513
+    },
+    {
+      "epoch": 8.8224,
+      "grad_norm": 0.0,
+      "learning_rate": 7.95e-05,
+      "loss": 2.5197,
+      "step": 5514
+    },
+    {
+      "epoch": 8.824,
+      "grad_norm": 0.0,
+      "learning_rate": 7.934999999999999e-05,
+      "loss": 2.7002,
+      "step": 5515
+    },
+    {
+      "epoch": 8.8256,
+      "grad_norm": 0.0,
+      "learning_rate": 7.92e-05,
+      "loss": 2.9525,
+      "step": 5516
+    },
+    {
+      "epoch": 8.8272,
+      "grad_norm": 0.0,
+      "learning_rate": 7.905e-05,
+      "loss": 3.031,
+      "step": 5517
+    },
+    {
+      "epoch": 8.8288,
+      "grad_norm": 0.0,
+      "learning_rate": 7.89e-05,
+      "loss": 2.4588,
+      "step": 5518
+    },
+    {
+      "epoch": 8.830400000000001,
+      "grad_norm": 0.0,
+      "learning_rate": 7.874999999999999e-05,
+      "loss": 3.0163,
+      "step": 5519
+    },
+    {
+      "epoch": 8.832,
+      "grad_norm": 0.0,
+      "learning_rate": 7.86e-05,
+      "loss": 2.6547,
+      "step": 5520
+    },
+    {
+      "epoch": 8.8336,
+      "grad_norm": 0.0,
+      "learning_rate": 7.845e-05,
+      "loss": 2.6753,
+      "step": 5521
+    },
+    {
+      "epoch": 8.8352,
+      "grad_norm": 0.0,
+      "learning_rate": 7.829999999999999e-05,
+      "loss": 2.4913,
+      "step": 5522
+    },
+    {
+      "epoch": 8.8368,
+      "grad_norm": 0.0,
+      "learning_rate": 7.815e-05,
+      "loss": 2.6377,
+      "step": 5523
+    },
+    {
+      "epoch": 8.8384,
+      "grad_norm": 0.0,
+      "learning_rate": 7.8e-05,
+      "loss": 2.8134,
+      "step": 5524
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 7.785e-05,
+      "loss": 2.7685,
+      "step": 5525
+    },
+    {
+      "epoch": 8.8416,
+      "grad_norm": 0.0,
+      "learning_rate": 7.769999999999999e-05,
+      "loss": 2.4115,
+      "step": 5526
+    },
+    {
+      "epoch": 8.8432,
+      "grad_norm": 0.0,
+      "learning_rate": 7.755e-05,
+      "loss": 3.4808,
+      "step": 5527
+    },
+    {
+      "epoch": 8.8448,
+      "grad_norm": 0.0,
+      "learning_rate": 7.74e-05,
+      "loss": 2.845,
+      "step": 5528
+    },
+    {
+      "epoch": 8.8464,
+      "grad_norm": 0.0,
+      "learning_rate": 7.725e-05,
+      "loss": 3.4166,
+      "step": 5529
+    },
+    {
+      "epoch": 8.848,
+      "grad_norm": 0.0,
+      "learning_rate": 7.709999999999999e-05,
+      "loss": 3.1478,
+      "step": 5530
+    },
+    {
+      "epoch": 8.8496,
+      "grad_norm": 0.0,
+      "learning_rate": 7.695e-05,
+      "loss": 3.3035,
+      "step": 5531
+    },
+    {
+      "epoch": 8.8512,
+      "grad_norm": 0.0,
+      "learning_rate": 7.68e-05,
+      "loss": 3.0173,
+      "step": 5532
+    },
+    {
+      "epoch": 8.8528,
+      "grad_norm": 0.0,
+      "learning_rate": 7.664999999999999e-05,
+      "loss": 3.3192,
+      "step": 5533
+    },
+    {
+      "epoch": 8.8544,
+      "grad_norm": 0.0,
+      "learning_rate": 7.649999999999999e-05,
+      "loss": 2.6861,
+      "step": 5534
+    },
+    {
+      "epoch": 8.856,
+      "grad_norm": 0.0,
+      "learning_rate": 7.635e-05,
+      "loss": 2.8469,
+      "step": 5535
+    },
+    {
+      "epoch": 8.8576,
+      "grad_norm": 0.0,
+      "learning_rate": 7.62e-05,
+      "loss": 2.5906,
+      "step": 5536
+    },
+    {
+      "epoch": 8.8592,
+      "grad_norm": 0.0,
+      "learning_rate": 7.604999999999999e-05,
+      "loss": 3.5129,
+      "step": 5537
+    },
+    {
+      "epoch": 8.8608,
+      "grad_norm": 0.0,
+      "learning_rate": 7.589999999999999e-05,
+      "loss": 2.6874,
+      "step": 5538
+    },
+    {
+      "epoch": 8.862400000000001,
+      "grad_norm": 0.0,
+      "learning_rate": 7.575e-05,
+      "loss": 2.8809,
+      "step": 5539
+    },
+    {
+      "epoch": 8.864,
+      "grad_norm": 0.0,
+      "learning_rate": 7.56e-05,
+      "loss": 2.9368,
+      "step": 5540
+    },
+    {
+      "epoch": 8.8656,
+      "grad_norm": 0.0,
+      "learning_rate": 7.544999999999999e-05,
+      "loss": 3.2077,
+      "step": 5541
+    },
+    {
+      "epoch": 8.8672,
+      "grad_norm": 0.0,
+      "learning_rate": 7.529999999999999e-05,
+      "loss": 2.355,
+      "step": 5542
+    },
+    {
+      "epoch": 8.8688,
+      "grad_norm": 0.0,
+      "learning_rate": 7.515e-05,
+      "loss": 2.6583,
+      "step": 5543
+    },
+    {
+      "epoch": 8.8704,
+      "grad_norm": 0.0,
+      "learning_rate": 7.5e-05,
+      "loss": 2.9874,
+      "step": 5544
+    },
+    {
+      "epoch": 8.872,
+      "grad_norm": 0.0,
+      "learning_rate": 7.484999999999999e-05,
+      "loss": 3.6936,
+      "step": 5545
+    },
+    {
+      "epoch": 8.8736,
+      "grad_norm": 0.0,
+      "learning_rate": 7.47e-05,
+      "loss": 3.0383,
+      "step": 5546
+    },
+    {
+      "epoch": 8.8752,
+      "grad_norm": 0.0,
+      "learning_rate": 7.455e-05,
+      "loss": 2.476,
+      "step": 5547
+    },
+    {
+      "epoch": 8.8768,
+      "grad_norm": 0.0,
+      "learning_rate": 7.439999999999999e-05,
+      "loss": 2.962,
+      "step": 5548
+    },
+    {
+      "epoch": 8.8784,
+      "grad_norm": 0.0,
+      "learning_rate": 7.424999999999999e-05,
+      "loss": 2.8282,
+      "step": 5549
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 7.41e-05,
+      "loss": 4.6808,
+      "step": 5550
+    },
+    {
+      "epoch": 8.8816,
+      "grad_norm": 0.0,
+      "learning_rate": 7.395e-05,
+      "loss": 5.0056,
+      "step": 5551
+    },
+    {
+      "epoch": 8.8832,
+      "grad_norm": 0.0,
+      "learning_rate": 7.379999999999999e-05,
+      "loss": 3.5208,
+      "step": 5552
+    },
+    {
+      "epoch": 8.8848,
+      "grad_norm": 0.0,
+      "learning_rate": 7.364999999999999e-05,
+      "loss": 3.9444,
+      "step": 5553
+    },
+    {
+      "epoch": 8.8864,
+      "grad_norm": 0.0,
+      "learning_rate": 7.35e-05,
+      "loss": 3.2421,
+      "step": 5554
+    },
+    {
+      "epoch": 8.888,
+      "grad_norm": 0.0,
+      "learning_rate": 7.335e-05,
+      "loss": 3.4752,
+      "step": 5555
+    },
+    {
+      "epoch": 8.8896,
+      "grad_norm": 0.0,
+      "learning_rate": 7.319999999999999e-05,
+      "loss": 2.616,
+      "step": 5556
+    },
+    {
+      "epoch": 8.8912,
+      "grad_norm": 0.0,
+      "learning_rate": 7.304999999999999e-05,
+      "loss": 2.9658,
+      "step": 5557
+    },
+    {
+      "epoch": 8.8928,
+      "grad_norm": 0.0,
+      "learning_rate": 7.29e-05,
+      "loss": 2.8145,
+      "step": 5558
+    },
+    {
+      "epoch": 8.8944,
+      "grad_norm": 0.0,
+      "learning_rate": 7.274999999999999e-05,
+      "loss": 2.8018,
+      "step": 5559
+    },
+    {
+      "epoch": 8.896,
+      "grad_norm": 0.0,
+      "learning_rate": 7.259999999999999e-05,
+      "loss": 3.2103,
+      "step": 5560
+    },
+    {
+      "epoch": 8.8976,
+      "grad_norm": 0.0,
+      "learning_rate": 7.244999999999999e-05,
+      "loss": 3.0906,
+      "step": 5561
+    },
+    {
+      "epoch": 8.8992,
+      "grad_norm": 0.0,
+      "learning_rate": 7.23e-05,
+      "loss": 2.5659,
+      "step": 5562
+    },
+    {
+      "epoch": 8.9008,
+      "grad_norm": 0.0,
+      "learning_rate": 7.214999999999999e-05,
+      "loss": 3.2138,
+      "step": 5563
+    },
+    {
+      "epoch": 8.9024,
+      "grad_norm": 0.0,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 2.8943,
+      "step": 5564
+    },
+    {
+      "epoch": 8.904,
+      "grad_norm": 0.0,
+      "learning_rate": 7.184999999999998e-05,
+      "loss": 2.9345,
+      "step": 5565
+    },
+    {
+      "epoch": 8.9056,
+      "grad_norm": 0.0,
+      "learning_rate": 7.17e-05,
+      "loss": 2.9651,
+      "step": 5566
+    },
+    {
+      "epoch": 8.9072,
+      "grad_norm": 0.0,
+      "learning_rate": 7.154999999999999e-05,
+      "loss": 3.0088,
+      "step": 5567
+    },
+    {
+      "epoch": 8.9088,
+      "grad_norm": 0.0,
+      "learning_rate": 7.139999999999999e-05,
+      "loss": 2.5948,
+      "step": 5568
+    },
+    {
+      "epoch": 8.9104,
+      "grad_norm": 0.0,
+      "learning_rate": 7.125e-05,
+      "loss": 2.509,
+      "step": 5569
+    },
+    {
+      "epoch": 8.912,
+      "grad_norm": 0.0,
+      "learning_rate": 7.11e-05,
+      "loss": 2.5297,
+      "step": 5570
+    },
+    {
+      "epoch": 8.9136,
+      "grad_norm": 0.0,
+      "learning_rate": 7.094999999999999e-05,
+      "loss": 2.6333,
+      "step": 5571
+    },
+    {
+      "epoch": 8.9152,
+      "grad_norm": 0.0,
+      "learning_rate": 7.079999999999999e-05,
+      "loss": 3.1458,
+      "step": 5572
+    },
+    {
+      "epoch": 8.9168,
+      "grad_norm": 0.0,
+      "learning_rate": 7.065e-05,
+      "loss": 2.8393,
+      "step": 5573
+    },
+    {
+      "epoch": 8.9184,
+      "grad_norm": 0.0,
+      "learning_rate": 7.049999999999999e-05,
+      "loss": 2.3593,
+      "step": 5574
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.034999999999999e-05,
+      "loss": 2.2004,
+      "step": 5575
+    },
+    {
+      "epoch": 8.9216,
+      "grad_norm": 0.0,
+      "learning_rate": 7.02e-05,
+      "loss": 3.1726,
+      "step": 5576
+    },
+    {
+      "epoch": 8.9232,
+      "grad_norm": 0.0,
+      "learning_rate": 7.005e-05,
+      "loss": 3.0657,
+      "step": 5577
+    },
+    {
+      "epoch": 8.9248,
+      "grad_norm": 0.0,
+      "learning_rate": 6.989999999999999e-05,
+      "loss": 2.9242,
+      "step": 5578
+    },
+    {
+      "epoch": 8.9264,
+      "grad_norm": 0.0,
+      "learning_rate": 6.975e-05,
+      "loss": 2.8571,
+      "step": 5579
+    },
+    {
+      "epoch": 8.928,
+      "grad_norm": 0.0,
+      "learning_rate": 6.96e-05,
+      "loss": 3.4707,
+      "step": 5580
+    },
+    {
+      "epoch": 8.9296,
+      "grad_norm": 0.0,
+      "learning_rate": 6.945e-05,
+      "loss": 2.372,
+      "step": 5581
+    },
+    {
+      "epoch": 8.9312,
+      "grad_norm": 0.0,
+      "learning_rate": 6.93e-05,
+      "loss": 2.6497,
+      "step": 5582
+    },
+    {
+      "epoch": 8.9328,
+      "grad_norm": 0.0,
+      "learning_rate": 6.915e-05,
+      "loss": 3.2633,
+      "step": 5583
+    },
+    {
+      "epoch": 8.9344,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9e-05,
+      "loss": 2.8765,
+      "step": 5584
+    },
+    {
+      "epoch": 8.936,
+      "grad_norm": 0.0,
+      "learning_rate": 6.884999999999999e-05,
+      "loss": 3.5971,
+      "step": 5585
+    },
+    {
+      "epoch": 8.9376,
+      "grad_norm": 0.0,
+      "learning_rate": 6.87e-05,
+      "loss": 3.4084,
+      "step": 5586
+    },
+    {
+      "epoch": 8.9392,
+      "grad_norm": 0.0,
+      "learning_rate": 6.855e-05,
+      "loss": 2.6225,
+      "step": 5587
+    },
+    {
+      "epoch": 8.9408,
+      "grad_norm": 0.0,
+      "learning_rate": 6.84e-05,
+      "loss": 2.7682,
+      "step": 5588
+    },
+    {
+      "epoch": 8.9424,
+      "grad_norm": 0.0,
+      "learning_rate": 6.824999999999999e-05,
+      "loss": 2.3212,
+      "step": 5589
+    },
+    {
+      "epoch": 8.943999999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 6.81e-05,
+      "loss": 2.308,
+      "step": 5590
+    },
+    {
+      "epoch": 8.9456,
+      "grad_norm": 0.0,
+      "learning_rate": 6.795e-05,
+      "loss": 3.1101,
+      "step": 5591
+    },
+    {
+      "epoch": 8.9472,
+      "grad_norm": 0.0,
+      "learning_rate": 6.78e-05,
+      "loss": 2.9434,
+      "step": 5592
+    },
+    {
+      "epoch": 8.9488,
+      "grad_norm": 0.0,
+      "learning_rate": 6.764999999999999e-05,
+      "loss": 4.3318,
+      "step": 5593
+    },
+    {
+      "epoch": 8.9504,
+      "grad_norm": 0.0,
+      "learning_rate": 6.75e-05,
+      "loss": 2.77,
+      "step": 5594
+    },
+    {
+      "epoch": 8.952,
+      "grad_norm": 0.0,
+      "learning_rate": 6.735e-05,
+      "loss": 2.4421,
+      "step": 5595
+    },
+    {
+      "epoch": 8.9536,
+      "grad_norm": 0.0,
+      "learning_rate": 6.72e-05,
+      "loss": 3.7881,
+      "step": 5596
+    },
+    {
+      "epoch": 8.9552,
+      "grad_norm": 0.0,
+      "learning_rate": 6.704999999999999e-05,
+      "loss": 2.5785,
+      "step": 5597
+    },
+    {
+      "epoch": 8.9568,
+      "grad_norm": 0.0,
+      "learning_rate": 6.69e-05,
+      "loss": 3.3298,
+      "step": 5598
+    },
+    {
+      "epoch": 8.9584,
+      "grad_norm": 0.0,
+      "learning_rate": 6.675e-05,
+      "loss": 3.5415,
+      "step": 5599
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.659999999999999e-05,
+      "loss": 4.0338,
+      "step": 5600
+    },
+    {
+      "epoch": 8.9616,
+      "grad_norm": 0.0,
+      "learning_rate": 6.644999999999999e-05,
+      "loss": 4.8242,
+      "step": 5601
+    },
+    {
+      "epoch": 8.9632,
+      "grad_norm": 0.0,
+      "learning_rate": 6.63e-05,
+      "loss": 4.5149,
+      "step": 5602
+    },
+    {
+      "epoch": 8.9648,
+      "grad_norm": 0.0,
+      "learning_rate": 6.615e-05,
+      "loss": 3.0374,
+      "step": 5603
+    },
+    {
+      "epoch": 8.9664,
+      "grad_norm": 0.0,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 3.8995,
+      "step": 5604
+    },
+    {
+      "epoch": 8.968,
+      "grad_norm": 0.0,
+      "learning_rate": 6.584999999999999e-05,
+      "loss": 2.7828,
+      "step": 5605
+    },
+    {
+      "epoch": 8.9696,
+      "grad_norm": 0.0,
+      "learning_rate": 6.57e-05,
+      "loss": 2.8965,
+      "step": 5606
+    },
+    {
+      "epoch": 8.9712,
+      "grad_norm": 0.0,
+      "learning_rate": 6.555e-05,
+      "loss": 2.3991,
+      "step": 5607
+    },
+    {
+      "epoch": 8.9728,
+      "grad_norm": 0.0,
+      "learning_rate": 6.539999999999999e-05,
+      "loss": 2.7713,
+      "step": 5608
+    },
+    {
+      "epoch": 8.9744,
+      "grad_norm": 0.0,
+      "learning_rate": 6.525e-05,
+      "loss": 4.3079,
+      "step": 5609
+    },
+    {
+      "epoch": 8.975999999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 6.51e-05,
+      "loss": 2.7852,
+      "step": 5610
+    },
+    {
+      "epoch": 8.9776,
+      "grad_norm": 0.0,
+      "learning_rate": 6.494999999999999e-05,
+      "loss": 2.3143,
+      "step": 5611
+    },
+    {
+      "epoch": 8.9792,
+      "grad_norm": 0.0,
+      "learning_rate": 6.479999999999999e-05,
+      "loss": 2.4507,
+      "step": 5612
+    },
+    {
+      "epoch": 8.9808,
+      "grad_norm": 0.0,
+      "learning_rate": 6.465e-05,
+      "loss": 2.6195,
+      "step": 5613
+    },
+    {
+      "epoch": 8.9824,
+      "grad_norm": 0.0,
+      "learning_rate": 6.45e-05,
+      "loss": 2.269,
+      "step": 5614
+    },
+    {
+      "epoch": 8.984,
+      "grad_norm": 0.0,
+      "learning_rate": 6.434999999999999e-05,
+      "loss": 3.2257,
+      "step": 5615
+    },
+    {
+      "epoch": 8.9856,
+      "grad_norm": 0.0,
+      "learning_rate": 6.419999999999999e-05,
+      "loss": 2.6905,
+      "step": 5616
+    },
+    {
+      "epoch": 8.9872,
+      "grad_norm": 0.0,
+      "learning_rate": 6.405e-05,
+      "loss": 2.8038,
+      "step": 5617
+    },
+    {
+      "epoch": 8.9888,
+      "grad_norm": 0.0,
+      "learning_rate": 6.39e-05,
+      "loss": 2.8442,
+      "step": 5618
+    },
+    {
+      "epoch": 8.9904,
+      "grad_norm": 0.0,
+      "learning_rate": 6.374999999999999e-05,
+      "loss": 2.72,
+      "step": 5619
+    },
+    {
+      "epoch": 8.992,
+      "grad_norm": 0.0,
+      "learning_rate": 6.359999999999999e-05,
+      "loss": 2.7504,
+      "step": 5620
+    },
+    {
+      "epoch": 8.9936,
+      "grad_norm": 0.0,
+      "learning_rate": 6.345e-05,
+      "loss": 3.0555,
+      "step": 5621
+    },
+    {
+      "epoch": 8.9952,
+      "grad_norm": 0.0,
+      "learning_rate": 6.33e-05,
+      "loss": 2.982,
+      "step": 5622
+    },
+    {
+      "epoch": 8.9968,
+      "grad_norm": 0.0,
+      "learning_rate": 6.314999999999999e-05,
+      "loss": 2.7094,
+      "step": 5623
+    },
+    {
+      "epoch": 8.9984,
+      "grad_norm": 0.0,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 3.1366,
+      "step": 5624
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.285e-05,
+      "loss": 3.7605,
+      "step": 5625
+    },
+    {
+      "epoch": 9.0016,
+      "grad_norm": 0.0,
+      "learning_rate": 6.269999999999999e-05,
+      "loss": 3.9877,
+      "step": 5626
+    },
+    {
+      "epoch": 9.0032,
+      "grad_norm": 0.0,
+      "learning_rate": 6.254999999999999e-05,
+      "loss": 4.1149,
+      "step": 5627
+    },
+    {
+      "epoch": 9.0048,
+      "grad_norm": 0.0,
+      "learning_rate": 6.239999999999999e-05,
+      "loss": 3.6217,
+      "step": 5628
+    },
+    {
+      "epoch": 9.0064,
+      "grad_norm": 0.0,
+      "learning_rate": 6.225e-05,
+      "loss": 4.6033,
+      "step": 5629
+    },
+    {
+      "epoch": 9.008,
+      "grad_norm": 0.0,
+      "learning_rate": 6.209999999999999e-05,
+      "loss": 3.7522,
+      "step": 5630
+    },
+    {
+      "epoch": 9.0096,
+      "grad_norm": 0.0,
+      "learning_rate": 6.194999999999999e-05,
+      "loss": 3.3814,
+      "step": 5631
+    },
+    {
+      "epoch": 9.0112,
+      "grad_norm": 0.0,
+      "learning_rate": 6.18e-05,
+      "loss": 2.7782,
+      "step": 5632
+    },
+    {
+      "epoch": 9.0128,
+      "grad_norm": 0.0,
+      "learning_rate": 6.165e-05,
+      "loss": 2.7994,
+      "step": 5633
+    },
+    {
+      "epoch": 9.0144,
+      "grad_norm": 0.0,
+      "learning_rate": 6.149999999999999e-05,
+      "loss": 2.7652,
+      "step": 5634
+    },
+    {
+      "epoch": 9.016,
+      "grad_norm": 0.0,
+      "learning_rate": 6.134999999999999e-05,
+      "loss": 2.9545,
+      "step": 5635
+    },
+    {
+      "epoch": 9.0176,
+      "grad_norm": 0.0,
+      "learning_rate": 6.12e-05,
+      "loss": 2.6188,
+      "step": 5636
+    },
+    {
+      "epoch": 9.0192,
+      "grad_norm": 0.0,
+      "learning_rate": 6.104999999999999e-05,
+      "loss": 3.047,
+      "step": 5637
+    },
+    {
+      "epoch": 9.0208,
+      "grad_norm": 0.0,
+      "learning_rate": 6.0899999999999996e-05,
+      "loss": 3.3657,
+      "step": 5638
+    },
+    {
+      "epoch": 9.0224,
+      "grad_norm": 0.0,
+      "learning_rate": 6.075e-05,
+      "loss": 3.0335,
+      "step": 5639
+    },
+    {
+      "epoch": 9.024,
+      "grad_norm": 0.0,
+      "learning_rate": 6.0599999999999996e-05,
+      "loss": 2.4625,
+      "step": 5640
+    },
+    {
+      "epoch": 9.0256,
+      "grad_norm": 0.0,
+      "learning_rate": 6.045e-05,
+      "loss": 2.4959,
+      "step": 5641
+    },
+    {
+      "epoch": 9.0272,
+      "grad_norm": 0.0,
+      "learning_rate": 6.0299999999999995e-05,
+      "loss": 3.5343,
+      "step": 5642
+    },
+    {
+      "epoch": 9.0288,
+      "grad_norm": 0.0,
+      "learning_rate": 6.015e-05,
+      "loss": 3.2499,
+      "step": 5643
+    },
+    {
+      "epoch": 9.0304,
+      "grad_norm": 0.0,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 3.5015,
+      "step": 5644
+    },
+    {
+      "epoch": 9.032,
+      "grad_norm": 0.0,
+      "learning_rate": 5.985e-05,
+      "loss": 2.3126,
+      "step": 5645
+    },
+    {
+      "epoch": 9.0336,
+      "grad_norm": 0.0,
+      "learning_rate": 5.97e-05,
+      "loss": 2.6735,
+      "step": 5646
+    },
+    {
+      "epoch": 9.0352,
+      "grad_norm": 0.0,
+      "learning_rate": 5.955e-05,
+      "loss": 2.4169,
+      "step": 5647
+    },
+    {
+      "epoch": 9.0368,
+      "grad_norm": 0.0,
+      "learning_rate": 5.94e-05,
+      "loss": 3.1192,
+      "step": 5648
+    },
+    {
+      "epoch": 9.0384,
+      "grad_norm": 0.0,
+      "learning_rate": 5.925e-05,
+      "loss": 3.5278,
+      "step": 5649
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.91e-05,
+      "loss": 3.8993,
+      "step": 5650
+    },
+    {
+      "epoch": 9.0416,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8949999999999996e-05,
+      "loss": 3.1883,
+      "step": 5651
+    },
+    {
+      "epoch": 9.0432,
+      "grad_norm": 0.0,
+      "learning_rate": 5.88e-05,
+      "loss": 2.9788,
+      "step": 5652
+    },
+    {
+      "epoch": 9.0448,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8649999999999996e-05,
+      "loss": 2.6266,
+      "step": 5653
+    },
+    {
+      "epoch": 9.0464,
+      "grad_norm": 0.0,
+      "learning_rate": 5.85e-05,
+      "loss": 2.5842,
+      "step": 5654
+    },
+    {
+      "epoch": 9.048,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8349999999999995e-05,
+      "loss": 2.5331,
+      "step": 5655
+    },
+    {
+      "epoch": 9.0496,
+      "grad_norm": 0.0,
+      "learning_rate": 5.82e-05,
+      "loss": 2.6838,
+      "step": 5656
+    },
+    {
+      "epoch": 9.0512,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8049999999999995e-05,
+      "loss": 2.9824,
+      "step": 5657
+    },
+    {
+      "epoch": 9.0528,
+      "grad_norm": 0.0,
+      "learning_rate": 5.79e-05,
+      "loss": 2.8042,
+      "step": 5658
+    },
+    {
+      "epoch": 9.0544,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7749999999999994e-05,
+      "loss": 2.4679,
+      "step": 5659
+    },
+    {
+      "epoch": 9.056,
+      "grad_norm": 0.0,
+      "learning_rate": 5.76e-05,
+      "loss": 2.2159,
+      "step": 5660
+    },
+    {
+      "epoch": 9.0576,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7449999999999994e-05,
+      "loss": 2.9483,
+      "step": 5661
+    },
+    {
+      "epoch": 9.0592,
+      "grad_norm": 0.0,
+      "learning_rate": 5.73e-05,
+      "loss": 3.2915,
+      "step": 5662
+    },
+    {
+      "epoch": 9.0608,
+      "grad_norm": 0.0,
+      "learning_rate": 5.714999999999999e-05,
+      "loss": 2.5169,
+      "step": 5663
+    },
+    {
+      "epoch": 9.0624,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 3.3614,
+      "step": 5664
+    },
+    {
+      "epoch": 9.064,
+      "grad_norm": 0.0,
+      "learning_rate": 5.684999999999999e-05,
+      "loss": 2.4081,
+      "step": 5665
+    },
+    {
+      "epoch": 9.0656,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6699999999999996e-05,
+      "loss": 3.4789,
+      "step": 5666
+    },
+    {
+      "epoch": 9.0672,
+      "grad_norm": 0.0,
+      "learning_rate": 5.654999999999999e-05,
+      "loss": 2.5808,
+      "step": 5667
+    },
+    {
+      "epoch": 9.0688,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 2.5403,
+      "step": 5668
+    },
+    {
+      "epoch": 9.0704,
+      "grad_norm": 0.0,
+      "learning_rate": 5.625e-05,
+      "loss": 2.92,
+      "step": 5669
+    },
+    {
+      "epoch": 9.072,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6099999999999995e-05,
+      "loss": 2.9666,
+      "step": 5670
+    },
+    {
+      "epoch": 9.0736,
+      "grad_norm": 0.0,
+      "learning_rate": 5.595e-05,
+      "loss": 3.6537,
+      "step": 5671
+    },
+    {
+      "epoch": 9.0752,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5799999999999994e-05,
+      "loss": 3.8247,
+      "step": 5672
+    },
+    {
+      "epoch": 9.0768,
+      "grad_norm": 0.0,
+      "learning_rate": 5.565e-05,
+      "loss": 3.4512,
+      "step": 5673
+    },
+    {
+      "epoch": 9.0784,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5499999999999994e-05,
+      "loss": 2.5999,
+      "step": 5674
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.535e-05,
+      "loss": 3.7474,
+      "step": 5675
+    },
+    {
+      "epoch": 9.0816,
+      "grad_norm": 0.0,
+      "learning_rate": 5.519999999999999e-05,
+      "loss": 3.7495,
+      "step": 5676
+    },
+    {
+      "epoch": 9.0832,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5049999999999996e-05,
+      "loss": 3.5728,
+      "step": 5677
+    },
+    {
+      "epoch": 9.0848,
+      "grad_norm": 0.0,
+      "learning_rate": 5.489999999999999e-05,
+      "loss": 3.3557,
+      "step": 5678
+    },
+    {
+      "epoch": 9.0864,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4749999999999996e-05,
+      "loss": 3.2979,
+      "step": 5679
+    },
+    {
+      "epoch": 9.088,
+      "grad_norm": 0.0,
+      "learning_rate": 5.459999999999999e-05,
+      "loss": 3.5679,
+      "step": 5680
+    },
+    {
+      "epoch": 9.0896,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4449999999999995e-05,
+      "loss": 2.8948,
+      "step": 5681
+    },
+    {
+      "epoch": 9.0912,
+      "grad_norm": 0.0,
+      "learning_rate": 5.429999999999999e-05,
+      "loss": 3.92,
+      "step": 5682
+    },
+    {
+      "epoch": 9.0928,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4149999999999995e-05,
+      "loss": 3.2817,
+      "step": 5683
+    },
+    {
+      "epoch": 9.0944,
+      "grad_norm": 0.0,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 3.2031,
+      "step": 5684
+    },
+    {
+      "epoch": 9.096,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3849999999999994e-05,
+      "loss": 2.8687,
+      "step": 5685
+    },
+    {
+      "epoch": 9.0976,
+      "grad_norm": 0.0,
+      "learning_rate": 5.369999999999999e-05,
+      "loss": 2.5963,
+      "step": 5686
+    },
+    {
+      "epoch": 9.0992,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3549999999999994e-05,
+      "loss": 2.6347,
+      "step": 5687
+    },
+    {
+      "epoch": 9.1008,
+      "grad_norm": 0.0,
+      "learning_rate": 5.339999999999999e-05,
+      "loss": 3.149,
+      "step": 5688
+    },
+    {
+      "epoch": 9.1024,
+      "grad_norm": 0.0,
+      "learning_rate": 5.324999999999999e-05,
+      "loss": 2.6927,
+      "step": 5689
+    },
+    {
+      "epoch": 9.104,
+      "grad_norm": 0.0,
+      "learning_rate": 5.309999999999999e-05,
+      "loss": 2.7637,
+      "step": 5690
+    },
+    {
+      "epoch": 9.1056,
+      "grad_norm": 0.0,
+      "learning_rate": 5.294999999999999e-05,
+      "loss": 2.6973,
+      "step": 5691
+    },
+    {
+      "epoch": 9.1072,
+      "grad_norm": 0.0,
+      "learning_rate": 5.279999999999999e-05,
+      "loss": 3.2617,
+      "step": 5692
+    },
+    {
+      "epoch": 9.1088,
+      "grad_norm": 0.0,
+      "learning_rate": 5.264999999999999e-05,
+      "loss": 3.2056,
+      "step": 5693
+    },
+    {
+      "epoch": 9.1104,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2499999999999995e-05,
+      "loss": 3.1844,
+      "step": 5694
+    },
+    {
+      "epoch": 9.112,
+      "grad_norm": 0.0,
+      "learning_rate": 5.234999999999999e-05,
+      "loss": 2.8712,
+      "step": 5695
+    },
+    {
+      "epoch": 9.1136,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2199999999999995e-05,
+      "loss": 2.5566,
+      "step": 5696
+    },
+    {
+      "epoch": 9.1152,
+      "grad_norm": 0.0,
+      "learning_rate": 5.204999999999999e-05,
+      "loss": 3.2583,
+      "step": 5697
+    },
+    {
+      "epoch": 9.1168,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1899999999999994e-05,
+      "loss": 3.2741,
+      "step": 5698
+    },
+    {
+      "epoch": 9.1184,
+      "grad_norm": 0.0,
+      "learning_rate": 5.174999999999999e-05,
+      "loss": 2.7779,
+      "step": 5699
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1599999999999994e-05,
+      "loss": 4.0854,
+      "step": 5700
+    },
+    {
+      "epoch": 9.1216,
+      "grad_norm": 0.0,
+      "learning_rate": 5.145e-05,
+      "loss": 3.047,
+      "step": 5701
+    },
+    {
+      "epoch": 9.1232,
+      "grad_norm": 0.0,
+      "learning_rate": 5.13e-05,
+      "loss": 2.4914,
+      "step": 5702
+    },
+    {
+      "epoch": 9.1248,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1149999999999996e-05,
+      "loss": 2.7614,
+      "step": 5703
+    },
+    {
+      "epoch": 9.1264,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1e-05,
+      "loss": 3.0712,
+      "step": 5704
+    },
+    {
+      "epoch": 9.128,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0849999999999996e-05,
+      "loss": 2.5899,
+      "step": 5705
+    },
+    {
+      "epoch": 9.1296,
+      "grad_norm": 0.0,
+      "learning_rate": 5.07e-05,
+      "loss": 3.2392,
+      "step": 5706
+    },
+    {
+      "epoch": 9.1312,
+      "grad_norm": 0.0,
+      "learning_rate": 5.055e-05,
+      "loss": 2.7351,
+      "step": 5707
+    },
+    {
+      "epoch": 9.1328,
+      "grad_norm": 0.0,
+      "learning_rate": 5.04e-05,
+      "loss": 3.9668,
+      "step": 5708
+    },
+    {
+      "epoch": 9.1344,
+      "grad_norm": 0.0,
+      "learning_rate": 5.025e-05,
+      "loss": 2.4666,
+      "step": 5709
+    },
+    {
+      "epoch": 9.136,
+      "grad_norm": 0.0,
+      "learning_rate": 5.01e-05,
+      "loss": 2.7733,
+      "step": 5710
+    },
+    {
+      "epoch": 9.1376,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995e-05,
+      "loss": 2.1372,
+      "step": 5711
+    },
+    {
+      "epoch": 9.1392,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98e-05,
+      "loss": 2.7009,
+      "step": 5712
+    },
+    {
+      "epoch": 9.1408,
+      "grad_norm": 0.0,
+      "learning_rate": 4.965e-05,
+      "loss": 2.9297,
+      "step": 5713
+    },
+    {
+      "epoch": 9.1424,
+      "grad_norm": 0.0,
+      "learning_rate": 4.95e-05,
+      "loss": 2.505,
+      "step": 5714
+    },
+    {
+      "epoch": 9.144,
+      "grad_norm": 0.0,
+      "learning_rate": 4.935e-05,
+      "loss": 2.9298,
+      "step": 5715
+    },
+    {
+      "epoch": 9.1456,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9199999999999997e-05,
+      "loss": 2.3865,
+      "step": 5716
+    },
+    {
+      "epoch": 9.1472,
+      "grad_norm": 0.0,
+      "learning_rate": 4.905e-05,
+      "loss": 2.2821,
+      "step": 5717
+    },
+    {
+      "epoch": 9.1488,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8899999999999996e-05,
+      "loss": 3.4511,
+      "step": 5718
+    },
+    {
+      "epoch": 9.1504,
+      "grad_norm": 0.0,
+      "learning_rate": 4.875e-05,
+      "loss": 3.0349,
+      "step": 5719
+    },
+    {
+      "epoch": 9.152,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8599999999999995e-05,
+      "loss": 3.9405,
+      "step": 5720
+    },
+    {
+      "epoch": 9.1536,
+      "grad_norm": 0.0,
+      "learning_rate": 4.845e-05,
+      "loss": 2.305,
+      "step": 5721
+    },
+    {
+      "epoch": 9.1552,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8299999999999995e-05,
+      "loss": 3.6042,
+      "step": 5722
+    },
+    {
+      "epoch": 9.1568,
+      "grad_norm": 0.0,
+      "learning_rate": 4.815e-05,
+      "loss": 3.039,
+      "step": 5723
+    },
+    {
+      "epoch": 9.1584,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 2.7975,
+      "step": 5724
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": NaN,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 2.1338,
+      "step": 5725
+    },
+    {
+      "epoch": 9.1616,
+      "grad_norm": 0.0,
+      "learning_rate": 4.785e-05,
+      "loss": 4.1658,
+      "step": 5726
+    },
+    {
+      "epoch": 9.1632,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7699999999999994e-05,
+      "loss": 3.3,
+      "step": 5727
+    },
+    {
+      "epoch": 9.1648,
+      "grad_norm": 0.0,
+      "learning_rate": 4.755e-05,
+      "loss": 3.6535,
+      "step": 5728
+    },
+    {
+      "epoch": 9.1664,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7399999999999993e-05,
+      "loss": 3.4712,
+      "step": 5729
+    },
+    {
+      "epoch": 9.168,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7249999999999997e-05,
+      "loss": 3.1756,
+      "step": 5730
+    },
+    {
+      "epoch": 9.1696,
+      "grad_norm": 0.0,
+      "learning_rate": 4.709999999999999e-05,
+      "loss": 2.7233,
+      "step": 5731
+    },
+    {
+      "epoch": 9.1712,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6949999999999996e-05,
+      "loss": 2.8601,
+      "step": 5732
+    },
+    {
+      "epoch": 9.1728,
+      "grad_norm": 0.0,
+      "learning_rate": 4.68e-05,
+      "loss": 4.1129,
+      "step": 5733
+    },
+    {
+      "epoch": 9.1744,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6649999999999996e-05,
+      "loss": 3.6857,
+      "step": 5734
+    },
+    {
+      "epoch": 9.176,
+      "grad_norm": 0.0,
+      "learning_rate": 4.65e-05,
+      "loss": 2.8752,
+      "step": 5735
+    },
+    {
+      "epoch": 9.1776,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6349999999999995e-05,
+      "loss": 2.7681,
+      "step": 5736
+    },
+    {
+      "epoch": 9.1792,
+      "grad_norm": 0.0,
+      "learning_rate": 4.62e-05,
+      "loss": 3.0197,
+      "step": 5737
+    },
+    {
+      "epoch": 9.1808,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6049999999999994e-05,
+      "loss": 2.6093,
+      "step": 5738
+    },
+    {
+      "epoch": 9.1824,
+      "grad_norm": 0.0,
+      "learning_rate": 4.59e-05,
+      "loss": 3.1584,
+      "step": 5739
+    },
+    {
+      "epoch": 9.184,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5749999999999994e-05,
+      "loss": 3.0305,
+      "step": 5740
+    },
+    {
+      "epoch": 9.1856,
+      "grad_norm": 0.0,
+      "learning_rate": 4.56e-05,
+      "loss": 2.4518,
+      "step": 5741
+    },
+    {
+      "epoch": 9.1872,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5449999999999993e-05,
+      "loss": 2.417,
+      "step": 5742
+    },
+    {
+      "epoch": 9.1888,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5299999999999997e-05,
+      "loss": 2.9896,
+      "step": 5743
+    },
+    {
+      "epoch": 9.1904,
+      "grad_norm": 0.0,
+      "learning_rate": 4.514999999999999e-05,
+      "loss": 3.043,
+      "step": 5744
+    },
+    {
+      "epoch": 9.192,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 2.8806,
+      "step": 5745
+    },
+    {
+      "epoch": 9.1936,
+      "grad_norm": 0.0,
+      "learning_rate": 4.484999999999999e-05,
+      "loss": 2.6109,
+      "step": 5746
+    },
+    {
+      "epoch": 9.1952,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4699999999999996e-05,
+      "loss": 2.5296,
+      "step": 5747
+    },
+    {
+      "epoch": 9.1968,
+      "grad_norm": 0.0,
+      "learning_rate": 4.454999999999999e-05,
+      "loss": 3.3942,
+      "step": 5748
+    },
+    {
+      "epoch": 9.1984,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4399999999999995e-05,
+      "loss": 2.9008,
+      "step": 5749
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.424999999999999e-05,
+      "loss": 3.6638,
+      "step": 5750
+    },
+    {
+      "epoch": 9.2016,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4099999999999995e-05,
+      "loss": 2.678,
+      "step": 5751
+    },
+    {
+      "epoch": 9.2032,
+      "grad_norm": 0.0,
+      "learning_rate": 4.394999999999999e-05,
+      "loss": 3.0434,
+      "step": 5752
+    },
+    {
+      "epoch": 9.2048,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3799999999999994e-05,
+      "loss": 3.5293,
+      "step": 5753
+    },
+    {
+      "epoch": 9.2064,
+      "grad_norm": 0.0,
+      "learning_rate": 4.364999999999999e-05,
+      "loss": 2.5569,
+      "step": 5754
+    },
+    {
+      "epoch": 9.208,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3499999999999993e-05,
+      "loss": 3.0115,
+      "step": 5755
+    },
+    {
+      "epoch": 9.2096,
+      "grad_norm": 0.0,
+      "learning_rate": 4.334999999999999e-05,
+      "loss": 2.4495,
+      "step": 5756
+    },
+    {
+      "epoch": 9.2112,
+      "grad_norm": 0.0,
+      "learning_rate": 4.319999999999999e-05,
+      "loss": 2.937,
+      "step": 5757
+    },
+    {
+      "epoch": 9.2128,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3049999999999996e-05,
+      "loss": 2.8026,
+      "step": 5758
+    },
+    {
+      "epoch": 9.2144,
+      "grad_norm": 0.0,
+      "learning_rate": 4.289999999999999e-05,
+      "loss": 2.8708,
+      "step": 5759
+    },
+    {
+      "epoch": 9.216,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2749999999999996e-05,
+      "loss": 3.5378,
+      "step": 5760
+    },
+    {
+      "epoch": 9.2176,
+      "grad_norm": 0.0,
+      "learning_rate": 4.259999999999999e-05,
+      "loss": 2.7786,
+      "step": 5761
+    },
+    {
+      "epoch": 9.2192,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2449999999999995e-05,
+      "loss": 2.6413,
+      "step": 5762
+    },
+    {
+      "epoch": 9.2208,
+      "grad_norm": 0.0,
+      "learning_rate": 4.229999999999999e-05,
+      "loss": 2.7484,
+      "step": 5763
+    },
+    {
+      "epoch": 9.2224,
+      "grad_norm": 0.0,
+      "learning_rate": 4.215e-05,
+      "loss": 2.9544,
+      "step": 5764
+    },
+    {
+      "epoch": 9.224,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2e-05,
+      "loss": 3.135,
+      "step": 5765
+    },
+    {
+      "epoch": 9.2256,
+      "grad_norm": 0.0,
+      "learning_rate": 4.185e-05,
+      "loss": 2.823,
+      "step": 5766
+    },
+    {
+      "epoch": 9.2272,
+      "grad_norm": 0.0,
+      "learning_rate": 4.17e-05,
+      "loss": 3.5126,
+      "step": 5767
+    },
+    {
+      "epoch": 9.2288,
+      "grad_norm": 0.0,
+      "learning_rate": 4.155e-05,
+      "loss": 3.7753,
+      "step": 5768
+    },
+    {
+      "epoch": 9.2304,
+      "grad_norm": 0.0,
+      "learning_rate": 4.14e-05,
+      "loss": 2.8611,
+      "step": 5769
+    },
+    {
+      "epoch": 9.232,
+      "grad_norm": 0.0,
+      "learning_rate": 4.125e-05,
+      "loss": 2.6387,
+      "step": 5770
+    },
+    {
+      "epoch": 9.2336,
+      "grad_norm": 0.0,
+      "learning_rate": 4.11e-05,
+      "loss": 2.8033,
+      "step": 5771
+    },
+    {
+      "epoch": 9.2352,
+      "grad_norm": 0.0,
+      "learning_rate": 4.095e-05,
+      "loss": 2.6439,
+      "step": 5772
+    },
+    {
+      "epoch": 9.2368,
+      "grad_norm": 0.0,
+      "learning_rate": 4.08e-05,
+      "loss": 2.7774,
+      "step": 5773
+    },
+    {
+      "epoch": 9.2384,
+      "grad_norm": 0.0,
+      "learning_rate": 4.065e-05,
+      "loss": 2.9947,
+      "step": 5774
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.05e-05,
+      "loss": 3.3206,
+      "step": 5775
+    },
+    {
+      "epoch": 9.2416,
+      "grad_norm": 0.0,
+      "learning_rate": 4.035e-05,
+      "loss": 3.8351,
+      "step": 5776
+    },
+    {
+      "epoch": 9.2432,
+      "grad_norm": 0.0,
+      "learning_rate": 4.02e-05,
+      "loss": 4.5679,
+      "step": 5777
+    },
+    {
+      "epoch": 9.2448,
+      "grad_norm": 0.0,
+      "learning_rate": 4.005e-05,
+      "loss": 3.8772,
+      "step": 5778
+    },
+    {
+      "epoch": 9.2464,
+      "grad_norm": 0.0,
+      "learning_rate": 3.99e-05,
+      "loss": 3.8999,
+      "step": 5779
+    },
+    {
+      "epoch": 9.248,
+      "grad_norm": 0.0,
+      "learning_rate": 3.975e-05,
+      "loss": 3.509,
+      "step": 5780
+    },
+    {
+      "epoch": 9.2496,
+      "grad_norm": 0.0,
+      "learning_rate": 3.96e-05,
+      "loss": 3.024,
+      "step": 5781
+    },
+    {
+      "epoch": 9.2512,
+      "grad_norm": 0.0,
+      "learning_rate": 3.945e-05,
+      "loss": 3.4285,
+      "step": 5782
+    },
+    {
+      "epoch": 9.2528,
+      "grad_norm": 0.0,
+      "learning_rate": 3.93e-05,
+      "loss": 3.8175,
+      "step": 5783
+    },
+    {
+      "epoch": 9.2544,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9149999999999996e-05,
+      "loss": 2.5193,
+      "step": 5784
+    },
+    {
+      "epoch": 9.256,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9e-05,
+      "loss": 3.9423,
+      "step": 5785
+    },
+    {
+      "epoch": 9.2576,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8849999999999996e-05,
+      "loss": 3.4668,
+      "step": 5786
+    },
+    {
+      "epoch": 9.2592,
+      "grad_norm": 0.0,
+      "learning_rate": 3.87e-05,
+      "loss": 3.8142,
+      "step": 5787
+    },
+    {
+      "epoch": 9.2608,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8549999999999995e-05,
+      "loss": 3.697,
+      "step": 5788
+    },
+    {
+      "epoch": 9.2624,
+      "grad_norm": 0.0,
+      "learning_rate": 3.84e-05,
+      "loss": 2.7373,
+      "step": 5789
+    },
+    {
+      "epoch": 9.264,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8249999999999995e-05,
+      "loss": 3.5196,
+      "step": 5790
+    },
+    {
+      "epoch": 9.2656,
+      "grad_norm": 0.0,
+      "learning_rate": 3.81e-05,
+      "loss": 2.7335,
+      "step": 5791
+    },
+    {
+      "epoch": 9.2672,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7949999999999994e-05,
+      "loss": 2.546,
+      "step": 5792
+    },
+    {
+      "epoch": 9.2688,
+      "grad_norm": 0.0,
+      "learning_rate": 3.78e-05,
+      "loss": 3.3374,
+      "step": 5793
+    },
+    {
+      "epoch": 9.2704,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7649999999999994e-05,
+      "loss": 2.9118,
+      "step": 5794
+    },
+    {
+      "epoch": 9.272,
+      "grad_norm": 0.0,
+      "learning_rate": 3.75e-05,
+      "loss": 3.0383,
+      "step": 5795
+    },
+    {
+      "epoch": 9.2736,
+      "grad_norm": 0.0,
+      "learning_rate": 3.735e-05,
+      "loss": 2.5258,
+      "step": 5796
+    },
+    {
+      "epoch": 9.2752,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7199999999999996e-05,
+      "loss": 2.2185,
+      "step": 5797
+    },
+    {
+      "epoch": 9.2768,
+      "grad_norm": 0.0,
+      "learning_rate": 3.705e-05,
+      "loss": 2.511,
+      "step": 5798
+    },
+    {
+      "epoch": 9.2784,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6899999999999996e-05,
+      "loss": 2.5834,
+      "step": 5799
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.675e-05,
+      "loss": 2.6032,
+      "step": 5800
+    },
+    {
+      "epoch": 9.2816,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6599999999999995e-05,
+      "loss": 2.407,
+      "step": 5801
+    },
+    {
+      "epoch": 9.2832,
+      "grad_norm": 0.0,
+      "learning_rate": 3.645e-05,
+      "loss": 2.821,
+      "step": 5802
+    },
+    {
+      "epoch": 9.2848,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6299999999999995e-05,
+      "loss": 2.4382,
+      "step": 5803
+    },
+    {
+      "epoch": 9.2864,
+      "grad_norm": 0.0,
+      "learning_rate": 3.615e-05,
+      "loss": 2.9931,
+      "step": 5804
+    },
+    {
+      "epoch": 9.288,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 2.8703,
+      "step": 5805
+    },
+    {
+      "epoch": 9.2896,
+      "grad_norm": 0.0,
+      "learning_rate": 3.585e-05,
+      "loss": 2.5434,
+      "step": 5806
+    },
+    {
+      "epoch": 9.2912,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5699999999999994e-05,
+      "loss": 3.2047,
+      "step": 5807
+    },
+    {
+      "epoch": 9.2928,
+      "grad_norm": 0.0,
+      "learning_rate": 3.555e-05,
+      "loss": 2.3115,
+      "step": 5808
+    },
+    {
+      "epoch": 9.2944,
+      "grad_norm": 0.0,
+      "learning_rate": 3.539999999999999e-05,
+      "loss": 2.7968,
+      "step": 5809
+    },
+    {
+      "epoch": 9.296,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5249999999999996e-05,
+      "loss": 3.1432,
+      "step": 5810
+    },
+    {
+      "epoch": 9.2976,
+      "grad_norm": 0.0,
+      "learning_rate": 3.51e-05,
+      "loss": 3.9086,
+      "step": 5811
+    },
+    {
+      "epoch": 9.2992,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4949999999999996e-05,
+      "loss": 2.4662,
+      "step": 5812
+    },
+    {
+      "epoch": 9.3008,
+      "grad_norm": 0.0,
+      "learning_rate": 3.48e-05,
+      "loss": 2.2656,
+      "step": 5813
+    },
+    {
+      "epoch": 9.3024,
+      "grad_norm": 0.0,
+      "learning_rate": 3.465e-05,
+      "loss": 2.8611,
+      "step": 5814
+    },
+    {
+      "epoch": 9.304,
+      "grad_norm": 0.0,
+      "learning_rate": 3.45e-05,
+      "loss": 3.6223,
+      "step": 5815
+    },
+    {
+      "epoch": 9.3056,
+      "grad_norm": 0.0,
+      "learning_rate": 3.435e-05,
+      "loss": 2.6435,
+      "step": 5816
+    },
+    {
+      "epoch": 9.3072,
+      "grad_norm": 0.0,
+      "learning_rate": 3.42e-05,
+      "loss": 3.2736,
+      "step": 5817
+    },
+    {
+      "epoch": 9.3088,
+      "grad_norm": 0.0,
+      "learning_rate": 3.405e-05,
+      "loss": 2.9351,
+      "step": 5818
+    },
+    {
+      "epoch": 9.3104,
+      "grad_norm": 0.0,
+      "learning_rate": 3.39e-05,
+      "loss": 2.6843,
+      "step": 5819
+    },
+    {
+      "epoch": 9.312,
+      "grad_norm": 0.0,
+      "learning_rate": 3.375e-05,
+      "loss": 2.9742,
+      "step": 5820
+    },
+    {
+      "epoch": 9.3136,
+      "grad_norm": 0.0,
+      "learning_rate": 3.36e-05,
+      "loss": 2.4212,
+      "step": 5821
+    },
+    {
+      "epoch": 9.3152,
+      "grad_norm": 0.0,
+      "learning_rate": 3.345e-05,
+      "loss": 3.1968,
+      "step": 5822
+    },
+    {
+      "epoch": 9.3168,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3299999999999996e-05,
+      "loss": 3.6791,
+      "step": 5823
+    },
+    {
+      "epoch": 9.3184,
+      "grad_norm": 0.0,
+      "learning_rate": 3.315e-05,
+      "loss": 2.5402,
+      "step": 5824
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 3.9026,
+      "step": 5825
+    },
+    {
+      "epoch": 9.3216,
+      "grad_norm": 0.0,
+      "learning_rate": 3.285e-05,
+      "loss": 3.9221,
+      "step": 5826
+    },
+    {
+      "epoch": 9.3232,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2699999999999995e-05,
+      "loss": 4.0257,
+      "step": 5827
+    },
+    {
+      "epoch": 9.3248,
+      "grad_norm": 0.0,
+      "learning_rate": 3.255e-05,
+      "loss": 4.8515,
+      "step": 5828
+    },
+    {
+      "epoch": 9.3264,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2399999999999995e-05,
+      "loss": 3.9743,
+      "step": 5829
+    },
+    {
+      "epoch": 9.328,
+      "grad_norm": 0.0,
+      "learning_rate": 3.225e-05,
+      "loss": 3.0707,
+      "step": 5830
+    },
+    {
+      "epoch": 9.3296,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2099999999999994e-05,
+      "loss": 3.5754,
+      "step": 5831
+    },
+    {
+      "epoch": 9.3312,
+      "grad_norm": 0.0,
+      "learning_rate": 3.195e-05,
+      "loss": 4.3465,
+      "step": 5832
+    },
+    {
+      "epoch": 9.3328,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1799999999999994e-05,
+      "loss": 3.3521,
+      "step": 5833
+    },
+    {
+      "epoch": 9.3344,
+      "grad_norm": 0.0,
+      "learning_rate": 3.165e-05,
+      "loss": 3.5843,
+      "step": 5834
+    },
+    {
+      "epoch": 9.336,
+      "grad_norm": 0.0,
+      "learning_rate": 3.149999999999999e-05,
+      "loss": 3.0415,
+      "step": 5835
+    },
+    {
+      "epoch": 9.3376,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1349999999999996e-05,
+      "loss": 2.9932,
+      "step": 5836
+    },
+    {
+      "epoch": 9.3392,
+      "grad_norm": 0.0,
+      "learning_rate": 3.119999999999999e-05,
+      "loss": 3.667,
+      "step": 5837
+    },
+    {
+      "epoch": 9.3408,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1049999999999996e-05,
+      "loss": 3.0368,
+      "step": 5838
+    },
+    {
+      "epoch": 9.3424,
+      "grad_norm": 0.0,
+      "learning_rate": 3.09e-05,
+      "loss": 3.2513,
+      "step": 5839
+    },
+    {
+      "epoch": 9.344,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0749999999999995e-05,
+      "loss": 2.686,
+      "step": 5840
+    },
+    {
+      "epoch": 9.3456,
+      "grad_norm": 0.0,
+      "learning_rate": 3.06e-05,
+      "loss": 4.3405,
+      "step": 5841
+    },
+    {
+      "epoch": 9.3472,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0449999999999998e-05,
+      "loss": 2.9989,
+      "step": 5842
+    },
+    {
+      "epoch": 9.3488,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0299999999999998e-05,
+      "loss": 2.8285,
+      "step": 5843
+    },
+    {
+      "epoch": 9.3504,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0149999999999998e-05,
+      "loss": 2.6507,
+      "step": 5844
+    },
+    {
+      "epoch": 9.352,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 2.6162,
+      "step": 5845
+    },
+    {
+      "epoch": 9.3536,
+      "grad_norm": 0.0,
+      "learning_rate": 2.985e-05,
+      "loss": 3.0373,
+      "step": 5846
+    },
+    {
+      "epoch": 9.3552,
+      "grad_norm": 0.0,
+      "learning_rate": 2.97e-05,
+      "loss": 2.9476,
+      "step": 5847
+    },
+    {
+      "epoch": 9.3568,
+      "grad_norm": 0.0,
+      "learning_rate": 2.955e-05,
+      "loss": 2.8315,
+      "step": 5848
+    },
+    {
+      "epoch": 9.3584,
+      "grad_norm": 0.0,
+      "learning_rate": 2.94e-05,
+      "loss": 2.36,
+      "step": 5849
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.925e-05,
+      "loss": 2.9085,
+      "step": 5850
+    },
+    {
+      "epoch": 9.3616,
+      "grad_norm": 0.0,
+      "learning_rate": 2.91e-05,
+      "loss": 2.8154,
+      "step": 5851
+    },
+    {
+      "epoch": 9.3632,
+      "grad_norm": 0.0,
+      "learning_rate": 2.895e-05,
+      "loss": 2.5466,
+      "step": 5852
+    },
+    {
+      "epoch": 9.3648,
+      "grad_norm": 0.0,
+      "learning_rate": 2.88e-05,
+      "loss": 2.9695,
+      "step": 5853
+    },
+    {
+      "epoch": 9.3664,
+      "grad_norm": 0.0,
+      "learning_rate": 2.865e-05,
+      "loss": 2.6365,
+      "step": 5854
+    },
+    {
+      "epoch": 9.368,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 2.7683,
+      "step": 5855
+    },
+    {
+      "epoch": 9.3696,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8349999999999998e-05,
+      "loss": 2.5878,
+      "step": 5856
+    },
+    {
+      "epoch": 9.3712,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 2.3464,
+      "step": 5857
+    },
+    {
+      "epoch": 9.3728,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8049999999999997e-05,
+      "loss": 2.5351,
+      "step": 5858
+    },
+    {
+      "epoch": 9.3744,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7899999999999997e-05,
+      "loss": 2.4149,
+      "step": 5859
+    },
+    {
+      "epoch": 9.376,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7749999999999997e-05,
+      "loss": 3.3166,
+      "step": 5860
+    },
+    {
+      "epoch": 9.3776,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7599999999999997e-05,
+      "loss": 4.1168,
+      "step": 5861
+    },
+    {
+      "epoch": 9.3792,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7449999999999996e-05,
+      "loss": 3.124,
+      "step": 5862
+    },
+    {
+      "epoch": 9.3808,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7299999999999996e-05,
+      "loss": 3.1329,
+      "step": 5863
+    },
+    {
+      "epoch": 9.3824,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7149999999999996e-05,
+      "loss": 3.0086,
+      "step": 5864
+    },
+    {
+      "epoch": 9.384,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 2.7413,
+      "step": 5865
+    },
+    {
+      "epoch": 9.3856,
+      "grad_norm": NaN,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 1.6222,
+      "step": 5866
+    },
+    {
+      "epoch": 9.3872,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6849999999999995e-05,
+      "loss": 2.4586,
+      "step": 5867
+    },
+    {
+      "epoch": 9.3888,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6699999999999995e-05,
+      "loss": 3.3739,
+      "step": 5868
+    },
+    {
+      "epoch": 9.3904,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6549999999999995e-05,
+      "loss": 2.6869,
+      "step": 5869
+    },
+    {
+      "epoch": 9.392,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6399999999999995e-05,
+      "loss": 3.0628,
+      "step": 5870
+    },
+    {
+      "epoch": 9.3936,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6249999999999998e-05,
+      "loss": 3.9845,
+      "step": 5871
+    },
+    {
+      "epoch": 9.395199999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6099999999999997e-05,
+      "loss": 2.5343,
+      "step": 5872
+    },
+    {
+      "epoch": 9.3968,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5949999999999997e-05,
+      "loss": 3.7584,
+      "step": 5873
+    },
+    {
+      "epoch": 9.3984,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5799999999999997e-05,
+      "loss": 2.9493,
+      "step": 5874
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": NaN,
+      "learning_rate": 2.5799999999999997e-05,
+      "loss": 3.2706,
+      "step": 5875
+    },
+    {
+      "epoch": 9.4016,
+      "grad_norm": 0.0,
+      "learning_rate": 2.565e-05,
+      "loss": 5.3678,
+      "step": 5876
+    },
+    {
+      "epoch": 9.4032,
+      "grad_norm": 0.0,
+      "learning_rate": 2.55e-05,
+      "loss": 3.8882,
+      "step": 5877
+    },
+    {
+      "epoch": 9.4048,
+      "grad_norm": 0.0,
+      "learning_rate": 2.535e-05,
+      "loss": 2.9439,
+      "step": 5878
+    },
+    {
+      "epoch": 9.4064,
+      "grad_norm": 0.0,
+      "learning_rate": 2.52e-05,
+      "loss": 2.6025,
+      "step": 5879
+    },
+    {
+      "epoch": 9.408,
+      "grad_norm": 0.0,
+      "learning_rate": 2.505e-05,
+      "loss": 3.2556,
+      "step": 5880
+    },
+    {
+      "epoch": 9.4096,
+      "grad_norm": 0.0,
+      "learning_rate": 2.49e-05,
+      "loss": 2.9982,
+      "step": 5881
+    },
+    {
+      "epoch": 9.411200000000001,
+      "grad_norm": 0.0,
+      "learning_rate": 2.475e-05,
+      "loss": 3.9662,
+      "step": 5882
+    },
+    {
+      "epoch": 9.4128,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 2.7577,
+      "step": 5883
+    },
+    {
+      "epoch": 9.4144,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4449999999999998e-05,
+      "loss": 4.4033,
+      "step": 5884
+    },
+    {
+      "epoch": 9.416,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4299999999999998e-05,
+      "loss": 2.8199,
+      "step": 5885
+    },
+    {
+      "epoch": 9.4176,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4149999999999997e-05,
+      "loss": 4.025,
+      "step": 5886
+    },
+    {
+      "epoch": 9.4192,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 3.2348,
+      "step": 5887
+    },
+    {
+      "epoch": 9.4208,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3849999999999997e-05,
+      "loss": 2.6776,
+      "step": 5888
+    },
+    {
+      "epoch": 9.4224,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3699999999999997e-05,
+      "loss": 3.5189,
+      "step": 5889
+    },
+    {
+      "epoch": 9.424,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3549999999999996e-05,
+      "loss": 2.7538,
+      "step": 5890
+    },
+    {
+      "epoch": 9.4256,
+      "grad_norm": 0.0,
+      "learning_rate": 2.34e-05,
+      "loss": 2.5793,
+      "step": 5891
+    },
+    {
+      "epoch": 9.4272,
+      "grad_norm": 0.0,
+      "learning_rate": 2.325e-05,
+      "loss": 3.0414,
+      "step": 5892
+    },
+    {
+      "epoch": 9.4288,
+      "grad_norm": 0.0,
+      "learning_rate": 2.31e-05,
+      "loss": 2.6425,
+      "step": 5893
+    },
+    {
+      "epoch": 9.4304,
+      "grad_norm": 0.0,
+      "learning_rate": 2.295e-05,
+      "loss": 2.6048,
+      "step": 5894
+    },
+    {
+      "epoch": 9.432,
+      "grad_norm": 0.0,
+      "learning_rate": 2.28e-05,
+      "loss": 2.5132,
+      "step": 5895
+    },
+    {
+      "epoch": 9.4336,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2649999999999998e-05,
+      "loss": 2.2403,
+      "step": 5896
+    },
+    {
+      "epoch": 9.4352,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2499999999999998e-05,
+      "loss": 3.1129,
+      "step": 5897
+    },
+    {
+      "epoch": 9.4368,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2349999999999998e-05,
+      "loss": 2.899,
+      "step": 5898
+    },
+    {
+      "epoch": 9.4384,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2199999999999998e-05,
+      "loss": 2.8279,
+      "step": 5899
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2049999999999997e-05,
+      "loss": 2.9215,
+      "step": 5900
+    },
+    {
+      "epoch": 9.4416,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1899999999999997e-05,
+      "loss": 2.9392,
+      "step": 5901
+    },
+    {
+      "epoch": 9.4432,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1749999999999997e-05,
+      "loss": 3.105,
+      "step": 5902
+    },
+    {
+      "epoch": 9.4448,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1599999999999996e-05,
+      "loss": 2.6916,
+      "step": 5903
+    },
+    {
+      "epoch": 9.4464,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1449999999999996e-05,
+      "loss": 3.3384,
+      "step": 5904
+    },
+    {
+      "epoch": 9.448,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1299999999999996e-05,
+      "loss": 2.9018,
+      "step": 5905
+    },
+    {
+      "epoch": 9.4496,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1149999999999996e-05,
+      "loss": 2.5655,
+      "step": 5906
+    },
+    {
+      "epoch": 9.4512,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1e-05,
+      "loss": 2.9827,
+      "step": 5907
+    },
+    {
+      "epoch": 9.4528,
+      "grad_norm": 0.0,
+      "learning_rate": 2.085e-05,
+      "loss": 3.165,
+      "step": 5908
+    },
+    {
+      "epoch": 9.4544,
+      "grad_norm": 0.0,
+      "learning_rate": 2.07e-05,
+      "loss": 2.5769,
+      "step": 5909
+    },
+    {
+      "epoch": 9.456,
+      "grad_norm": 0.0,
+      "learning_rate": 2.055e-05,
+      "loss": 3.0794,
+      "step": 5910
+    },
+    {
+      "epoch": 9.4576,
+      "grad_norm": 0.0,
+      "learning_rate": 2.04e-05,
+      "loss": 2.8733,
+      "step": 5911
+    },
+    {
+      "epoch": 9.4592,
+      "grad_norm": 0.0,
+      "learning_rate": 2.025e-05,
+      "loss": 2.8476,
+      "step": 5912
+    },
+    {
+      "epoch": 9.4608,
+      "grad_norm": 0.0,
+      "learning_rate": 2.01e-05,
+      "loss": 2.986,
+      "step": 5913
+    },
+    {
+      "epoch": 9.4624,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995e-05,
+      "loss": 2.6161,
+      "step": 5914
+    },
+    {
+      "epoch": 9.464,
+      "grad_norm": 0.0,
+      "learning_rate": 1.98e-05,
+      "loss": 2.8138,
+      "step": 5915
+    },
+    {
+      "epoch": 9.4656,
+      "grad_norm": 0.0,
+      "learning_rate": 1.965e-05,
+      "loss": 2.5781,
+      "step": 5916
+    },
+    {
+      "epoch": 9.4672,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95e-05,
+      "loss": 3.0813,
+      "step": 5917
+    },
+    {
+      "epoch": 9.4688,
+      "grad_norm": 0.0,
+      "learning_rate": 1.935e-05,
+      "loss": 2.7178,
+      "step": 5918
+    },
+    {
+      "epoch": 9.4704,
+      "grad_norm": 0.0,
+      "learning_rate": 1.92e-05,
+      "loss": 2.3634,
+      "step": 5919
+    },
+    {
+      "epoch": 9.472,
+      "grad_norm": 0.0,
+      "learning_rate": 1.905e-05,
+      "loss": 3.006,
+      "step": 5920
+    },
+    {
+      "epoch": 9.4736,
+      "grad_norm": 0.0,
+      "learning_rate": 1.89e-05,
+      "loss": 2.4351,
+      "step": 5921
+    },
+    {
+      "epoch": 9.4752,
+      "grad_norm": 0.0,
+      "learning_rate": 1.875e-05,
+      "loss": 2.6789,
+      "step": 5922
+    },
+    {
+      "epoch": 9.4768,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8599999999999998e-05,
+      "loss": 3.3793,
+      "step": 5923
+    },
+    {
+      "epoch": 9.4784,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8449999999999998e-05,
+      "loss": 3.5019,
+      "step": 5924
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8299999999999998e-05,
+      "loss": 3.5246,
+      "step": 5925
+    },
+    {
+      "epoch": 9.4816,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8149999999999997e-05,
+      "loss": 3.9433,
+      "step": 5926
+    },
+    {
+      "epoch": 9.4832,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 4.1305,
+      "step": 5927
+    },
+    {
+      "epoch": 9.4848,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7849999999999997e-05,
+      "loss": 4.3702,
+      "step": 5928
+    },
+    {
+      "epoch": 9.4864,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7699999999999997e-05,
+      "loss": 3.1913,
+      "step": 5929
+    },
+    {
+      "epoch": 9.488,
+      "grad_norm": 0.0,
+      "learning_rate": 1.755e-05,
+      "loss": 3.7054,
+      "step": 5930
+    },
+    {
+      "epoch": 9.4896,
+      "grad_norm": 0.0,
+      "learning_rate": 1.74e-05,
+      "loss": 3.1622,
+      "step": 5931
+    },
+    {
+      "epoch": 9.4912,
+      "grad_norm": 0.0,
+      "learning_rate": 1.725e-05,
+      "loss": 3.7106,
+      "step": 5932
+    },
+    {
+      "epoch": 9.4928,
+      "grad_norm": 0.0,
+      "learning_rate": 1.71e-05,
+      "loss": 3.2574,
+      "step": 5933
+    },
+    {
+      "epoch": 9.4944,
+      "grad_norm": 0.0,
+      "learning_rate": 1.695e-05,
+      "loss": 2.7068,
+      "step": 5934
+    },
+    {
+      "epoch": 9.496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.68e-05,
+      "loss": 3.2067,
+      "step": 5935
+    },
+    {
+      "epoch": 9.4976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6649999999999998e-05,
+      "loss": 2.554,
+      "step": 5936
+    },
+    {
+      "epoch": 9.4992,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6499999999999998e-05,
+      "loss": 2.9718,
+      "step": 5937
+    },
+    {
+      "epoch": 9.5008,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6349999999999998e-05,
+      "loss": 2.844,
+      "step": 5938
+    },
+    {
+      "epoch": 9.5024,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6199999999999997e-05,
+      "loss": 3.0721,
+      "step": 5939
+    },
+    {
+      "epoch": 9.504,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6049999999999997e-05,
+      "loss": 2.6736,
+      "step": 5940
+    },
+    {
+      "epoch": 9.5056,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5899999999999997e-05,
+      "loss": 2.76,
+      "step": 5941
+    },
+    {
+      "epoch": 9.5072,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5749999999999997e-05,
+      "loss": 2.7879,
+      "step": 5942
+    },
+    {
+      "epoch": 9.5088,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5599999999999996e-05,
+      "loss": 3.4875,
+      "step": 5943
+    },
+    {
+      "epoch": 9.5104,
+      "grad_norm": 0.0,
+      "learning_rate": 1.545e-05,
+      "loss": 2.5975,
+      "step": 5944
+    },
+    {
+      "epoch": 9.512,
+      "grad_norm": 0.0,
+      "learning_rate": 1.53e-05,
+      "loss": 2.9525,
+      "step": 5945
+    },
+    {
+      "epoch": 9.5136,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5149999999999999e-05,
+      "loss": 3.0019,
+      "step": 5946
+    },
+    {
+      "epoch": 9.5152,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 2.6932,
+      "step": 5947
+    },
+    {
+      "epoch": 9.5168,
+      "grad_norm": 0.0,
+      "learning_rate": 1.485e-05,
+      "loss": 2.3499,
+      "step": 5948
+    },
+    {
+      "epoch": 9.5184,
+      "grad_norm": 0.0,
+      "learning_rate": 1.47e-05,
+      "loss": 2.7315,
+      "step": 5949
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.455e-05,
+      "loss": 2.4128,
+      "step": 5950
+    },
+    {
+      "epoch": 9.5216,
+      "grad_norm": 0.0,
+      "learning_rate": 1.44e-05,
+      "loss": 3.1493,
+      "step": 5951
+    },
+    {
+      "epoch": 9.5232,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 3.5148,
+      "step": 5952
+    },
+    {
+      "epoch": 9.5248,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 2.7583,
+      "step": 5953
+    },
+    {
+      "epoch": 9.5264,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3949999999999999e-05,
+      "loss": 2.7706,
+      "step": 5954
+    },
+    {
+      "epoch": 9.528,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3799999999999998e-05,
+      "loss": 3.7789,
+      "step": 5955
+    },
+    {
+      "epoch": 9.5296,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3649999999999998e-05,
+      "loss": 3.1102,
+      "step": 5956
+    },
+    {
+      "epoch": 9.5312,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3499999999999998e-05,
+      "loss": 3.0082,
+      "step": 5957
+    },
+    {
+      "epoch": 9.5328,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3349999999999998e-05,
+      "loss": 2.8169,
+      "step": 5958
+    },
+    {
+      "epoch": 9.5344,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3199999999999997e-05,
+      "loss": 3.482,
+      "step": 5959
+    },
+    {
+      "epoch": 9.536,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3049999999999999e-05,
+      "loss": 3.1191,
+      "step": 5960
+    },
+    {
+      "epoch": 9.5376,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2899999999999998e-05,
+      "loss": 3.3154,
+      "step": 5961
+    },
+    {
+      "epoch": 9.5392,
+      "grad_norm": 0.0,
+      "learning_rate": 1.275e-05,
+      "loss": 3.1597,
+      "step": 5962
+    },
+    {
+      "epoch": 9.5408,
+      "grad_norm": 0.0,
+      "learning_rate": 1.26e-05,
+      "loss": 2.663,
+      "step": 5963
+    },
+    {
+      "epoch": 9.5424,
+      "grad_norm": 0.0,
+      "learning_rate": 1.245e-05,
+      "loss": 3.9024,
+      "step": 5964
+    },
+    {
+      "epoch": 9.544,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2299999999999999e-05,
+      "loss": 3.2608,
+      "step": 5965
+    },
+    {
+      "epoch": 9.5456,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2149999999999999e-05,
+      "loss": 2.6814,
+      "step": 5966
+    },
+    {
+      "epoch": 9.5472,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 3.5144,
+      "step": 5967
+    },
+    {
+      "epoch": 9.5488,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1849999999999998e-05,
+      "loss": 2.4768,
+      "step": 5968
+    },
+    {
+      "epoch": 9.5504,
+      "grad_norm": 0.0,
+      "learning_rate": 1.17e-05,
+      "loss": 2.67,
+      "step": 5969
+    },
+    {
+      "epoch": 9.552,
+      "grad_norm": 0.0,
+      "learning_rate": 1.155e-05,
+      "loss": 2.5965,
+      "step": 5970
+    },
+    {
+      "epoch": 9.5536,
+      "grad_norm": 0.0,
+      "learning_rate": 1.14e-05,
+      "loss": 2.9419,
+      "step": 5971
+    },
+    {
+      "epoch": 9.5552,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1249999999999999e-05,
+      "loss": 2.772,
+      "step": 5972
+    },
+    {
+      "epoch": 9.556799999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1099999999999999e-05,
+      "loss": 3.3538,
+      "step": 5973
+    },
+    {
+      "epoch": 9.5584,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0949999999999998e-05,
+      "loss": 3.2122,
+      "step": 5974
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0799999999999998e-05,
+      "loss": 2.8816,
+      "step": 5975
+    },
+    {
+      "epoch": 9.5616,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0649999999999998e-05,
+      "loss": 3.9349,
+      "step": 5976
+    },
+    {
+      "epoch": 9.5632,
+      "grad_norm": 0.0,
+      "learning_rate": 1.05e-05,
+      "loss": 3.7896,
+      "step": 5977
+    },
+    {
+      "epoch": 9.5648,
+      "grad_norm": 0.0,
+      "learning_rate": 1.035e-05,
+      "loss": 3.084,
+      "step": 5978
+    },
+    {
+      "epoch": 9.5664,
+      "grad_norm": 0.0,
+      "learning_rate": 1.02e-05,
+      "loss": 4.2844,
+      "step": 5979
+    },
+    {
+      "epoch": 9.568,
+      "grad_norm": 0.0,
+      "learning_rate": 1.005e-05,
+      "loss": 3.3977,
+      "step": 5980
+    },
+    {
+      "epoch": 9.5696,
+      "grad_norm": 0.0,
+      "learning_rate": 9.9e-06,
+      "loss": 3.6734,
+      "step": 5981
+    },
+    {
+      "epoch": 9.5712,
+      "grad_norm": 0.0,
+      "learning_rate": 9.75e-06,
+      "loss": 3.9535,
+      "step": 5982
+    },
+    {
+      "epoch": 9.5728,
+      "grad_norm": 0.0,
+      "learning_rate": 9.6e-06,
+      "loss": 2.9579,
+      "step": 5983
+    },
+    {
+      "epoch": 9.5744,
+      "grad_norm": 0.0,
+      "learning_rate": 9.45e-06,
+      "loss": 3.5256,
+      "step": 5984
+    },
+    {
+      "epoch": 9.576,
+      "grad_norm": 0.0,
+      "learning_rate": 9.299999999999999e-06,
+      "loss": 2.9872,
+      "step": 5985
+    },
+    {
+      "epoch": 9.5776,
+      "grad_norm": 0.0,
+      "learning_rate": 9.149999999999999e-06,
+      "loss": 2.7637,
+      "step": 5986
+    },
+    {
+      "epoch": 9.5792,
+      "grad_norm": 0.0,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 2.3799,
+      "step": 5987
+    },
+    {
+      "epoch": 9.5808,
+      "grad_norm": 0.0,
+      "learning_rate": 8.849999999999998e-06,
+      "loss": 2.8928,
+      "step": 5988
+    },
+    {
+      "epoch": 9.5824,
+      "grad_norm": 0.0,
+      "learning_rate": 8.7e-06,
+      "loss": 2.941,
+      "step": 5989
+    },
+    {
+      "epoch": 9.584,
+      "grad_norm": 0.0,
+      "learning_rate": 8.55e-06,
+      "loss": 2.6913,
+      "step": 5990
+    },
+    {
+      "epoch": 9.5856,
+      "grad_norm": 0.0,
+      "learning_rate": 8.4e-06,
+      "loss": 2.8565,
+      "step": 5991
+    },
+    {
+      "epoch": 9.5872,
+      "grad_norm": 0.0,
+      "learning_rate": 8.249999999999999e-06,
+      "loss": 2.6545,
+      "step": 5992
+    },
+    {
+      "epoch": 9.588799999999999,
+      "grad_norm": 0.0,
+      "learning_rate": 8.099999999999999e-06,
+      "loss": 2.7364,
+      "step": 5993
+    },
+    {
+      "epoch": 9.5904,
+      "grad_norm": 0.0,
+      "learning_rate": 7.949999999999998e-06,
+      "loss": 2.4076,
+      "step": 5994
+    },
+    {
+      "epoch": 9.592,
+      "grad_norm": 0.0,
+      "learning_rate": 7.799999999999998e-06,
+      "loss": 2.8067,
+      "step": 5995
+    },
+    {
+      "epoch": 9.5936,
+      "grad_norm": 0.0,
+      "learning_rate": 7.65e-06,
+      "loss": 2.4151,
+      "step": 5996
+    },
+    {
+      "epoch": 9.5952,
+      "grad_norm": 0.0,
+      "learning_rate": 7.499999999999999e-06,
+      "loss": 2.9682,
+      "step": 5997
+    },
+    {
+      "epoch": 9.5968,
+      "grad_norm": 0.0,
+      "learning_rate": 7.35e-06,
+      "loss": 2.9539,
+      "step": 5998
+    },
+    {
+      "epoch": 9.5984,
+      "grad_norm": 0.0,
+      "learning_rate": 7.2e-06,
+      "loss": 3.2376,
+      "step": 5999
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.0,
+      "learning_rate": 7.049999999999999e-06,
+      "loss": 3.1155,
+      "step": 6000
+    },
+    {
+      "epoch": 9.6,
+      "eval_cer": 0.5195918204647826,
+      "eval_loss": 3.2885050773620605,
+      "eval_runtime": 157.437,
+      "eval_samples_per_second": 19.919,
+      "eval_steps_per_second": 1.245,
+      "eval_wer": 0.8050150243498083,
+      "step": 6000
+    },
+    {
+      "epoch": 9.6,
+      "step": 6000,
+      "total_flos": 3.700768773245485e+19,
+      "train_loss": 2.825458660195271,
+      "train_runtime": 12931.7591,
+      "train_samples_per_second": 14.847,
+      "train_steps_per_second": 0.464
     }
   ],
   "logging_steps": 1.0,
-  "max_steps": 3500,
+  "max_steps": 6000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 6,
-  "save_steps": 500,
+  "num_train_epochs": 10,
+  "save_steps": 1000,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -24885,7 +42095,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.158464150901847e+19,
+  "total_flos": 3.700768773245485e+19,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null