diff --git "a/checkpoint-534240/trainer_state.json" "b/checkpoint-534240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-534240/trainer_state.json"
@@ -0,0 +1,6339 @@
+{
+  "best_metric": 4.150570869445801,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/lstm/3/checkpoints/checkpoint-534240",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 534240,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8193,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5528,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0655,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 7.0038,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9567,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9333,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7583,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6494,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5522,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4697,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.4108,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3505,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.279,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.2115,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.163,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.1099,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.064,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.0229,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9858,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9451,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.9127,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8723,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8504,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.8095,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7896,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7592,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7363,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.7164,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6864,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.661,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6484,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6377,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.6148,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5928,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5877,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5604,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5467,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.5138,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.517,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.495,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4865,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4727,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4673,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.4368,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.4292,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.4252,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.4094,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.4115,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3897,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3689,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3782,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3554,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3494,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.3231,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.3342,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.3097,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.337,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2988,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2995,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2921,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2727,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2601,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2592,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 5.2427,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.2365,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2539,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.2348,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.2196,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1965,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1937,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.2021,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.199,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1908,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1877,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1795,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.172,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1582,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1481,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.143,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.1294,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1368,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.1283,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.1304,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.1112,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0917,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.106,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.1017,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0989,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0846,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0866,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0716,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0666,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0659,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0585,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0629,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0555,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.054,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 5.0308,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0282,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 5.0263,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 5.0243,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 5.0169,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 5.0106,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9992,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 5.0068,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9826,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9882,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9854,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9772,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.972,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9762,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.963,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.9663,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052682749778396e-05,
+      "loss": 4.9583,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9476,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9528,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9511,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.9449,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.9299,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.936,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.937,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.9266,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.9132,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.9205,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.9208,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.9077,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.9024,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8932,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8968,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.887,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.9026,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8919,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8734,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.88,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8784,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.874,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8784,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8631,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8654,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8722,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8551,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8518,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8525,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8396,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879281664977466e-05,
+      "loss": 4.8387,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878443070226414e-05,
+      "loss": 4.8447,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877604475475362e-05,
+      "loss": 4.8284,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87676588072431e-05,
+      "loss": 4.8317,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.8373,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.8327,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.789715766906738,
+      "eval_runtime": 296.8384,
+      "eval_samples_per_second": 1285.518,
+      "eval_steps_per_second": 40.173,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.8151,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.8145,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.8291,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.8121,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.8245,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87006039847664e-05,
+      "loss": 4.7966,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.8152,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7978,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7922,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7973,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.8009,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650304678507016e-05,
+      "loss": 4.8013,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8641935109800225e-05,
+      "loss": 4.7841,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633549162289705e-05,
+      "loss": 4.7819,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625163214779185e-05,
+      "loss": 4.7754,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616777267268665e-05,
+      "loss": 4.7762,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.774,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7742,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.7691,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.7882,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.7655,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.7741,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558092013498754e-05,
+      "loss": 4.761,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7688,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.7565,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.7563,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.751,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851617865474989e-05,
+      "loss": 4.7469,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850779270723937e-05,
+      "loss": 4.7304,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849940675972885e-05,
+      "loss": 4.7463,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849102081221833e-05,
+      "loss": 4.7427,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848265124351154e-05,
+      "loss": 4.7461,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847426529600102e-05,
+      "loss": 4.7357,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.7385,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.7318,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.7271,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.7123,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.7198,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.7173,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.7149,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.7182,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839880814721007e-05,
+      "loss": 4.7172,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.7068,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838205263099276e-05,
+      "loss": 4.712,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373666683482236e-05,
+      "loss": 4.7121,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365280735971716e-05,
+      "loss": 4.6985,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356894788461196e-05,
+      "loss": 4.7099,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6964,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6847,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.7038,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832338375602659e-05,
+      "loss": 4.6868,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831499780851606e-05,
+      "loss": 4.6861,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830661186100554e-05,
+      "loss": 4.6745,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829822591349502e-05,
+      "loss": 4.6848,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82898399659845e-05,
+      "loss": 4.6708,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828145401847398e-05,
+      "loss": 4.7047,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827306807096346e-05,
+      "loss": 4.6729,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826468212345294e-05,
+      "loss": 4.6864,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.6799,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6672,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6574,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6714,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6572,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6562,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6737,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.6682,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189241353465726e-05,
+      "loss": 4.6522,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180871784758935e-05,
+      "loss": 4.6456,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172485837248415e-05,
+      "loss": 4.646,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164099889737895e-05,
+      "loss": 4.6578,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155713942227375e-05,
+      "loss": 4.6571,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147327994716854e-05,
+      "loss": 4.6546,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.6547,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.6546,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6548,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6374,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6412,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.6403,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.6251,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.6399,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.6376,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6405,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.6218,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804676213980554e-05,
+      "loss": 4.6123,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803837619229502e-05,
+      "loss": 4.6298,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8029990244784504e-05,
+      "loss": 4.624,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021604297273984e-05,
+      "loss": 4.6306,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013218349763464e-05,
+      "loss": 4.6225,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004832402252944e-05,
+      "loss": 4.6237,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996446454742424e-05,
+      "loss": 4.6187,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79880605072319e-05,
+      "loss": 4.6113,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797967455972138e-05,
+      "loss": 4.6174,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797128861221086e-05,
+      "loss": 4.613,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.6213,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954533095993546e-05,
+      "loss": 4.6214,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946147148483026e-05,
+      "loss": 4.6173,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937761200972506e-05,
+      "loss": 4.6048,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929375253461986e-05,
+      "loss": 4.6026,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.6014,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791261973724468e-05,
+      "loss": 4.6118,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790423378973416e-05,
+      "loss": 4.5968,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789584784222364e-05,
+      "loss": 4.5973,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788746189471312e-05,
+      "loss": 4.5909,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78790759472026e-05,
+      "loss": 4.6058,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787070637849581e-05,
+      "loss": 4.5785,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786232043098529e-05,
+      "loss": 4.5917,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78539508622785e-05,
+      "loss": 4.5929,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784556491476798e-05,
+      "loss": 4.5862,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783717896725746e-05,
+      "loss": 4.5834,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.5951,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782040707223642e-05,
+      "loss": 4.5796,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812021124725906e-05,
+      "loss": 4.5825,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803635177215386e-05,
+      "loss": 4.5873,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795249229704866e-05,
+      "loss": 4.5734,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786879660998075e-05,
+      "loss": 4.5847,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.5881,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770124144780764e-05,
+      "loss": 4.5806,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761738197270244e-05,
+      "loss": 4.5692,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753352249759724e-05,
+      "loss": 4.5789,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744966302249204e-05,
+      "loss": 4.5791,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736580354738684e-05,
+      "loss": 4.5702,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728194407228164e-05,
+      "loss": 4.5644,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771982483852137e-05,
+      "loss": 4.569,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771143889101086e-05,
+      "loss": 4.5791,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770305294350034e-05,
+      "loss": 4.5664,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769466699598982e-05,
+      "loss": 4.564,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76862810484793e-05,
+      "loss": 4.5531,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5607,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.5547,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766113958475147e-05,
+      "loss": 4.572,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765275363724095e-05,
+      "loss": 4.5675,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764436768973043e-05,
+      "loss": 4.5468,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76359817422199e-05,
+      "loss": 4.5544,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762759579470938e-05,
+      "loss": 4.5588,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.5555,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5632,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760245433098156e-05,
+      "loss": 4.5486,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594084762274773e-05,
+      "loss": 4.551,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758569881476425e-05,
+      "loss": 4.5585,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757731286725373e-05,
+      "loss": 4.5456,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5474,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.5528,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75521714035259e-05,
+      "loss": 4.5369,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543785456015375e-05,
+      "loss": 4.5411,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.548,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.5321,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5397,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510241665973295e-05,
+      "loss": 4.5431,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750185571846278e-05,
+      "loss": 4.5393,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.501766681671143,
+      "eval_runtime": 297.3924,
+      "eval_samples_per_second": 1283.123,
+      "eval_steps_per_second": 40.099,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749346977095226e-05,
+      "loss": 4.5264,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.528,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.5475,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5259,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5412,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.521,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.5358,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743476813837862e-05,
+      "loss": 4.5181,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426382190868095e-05,
+      "loss": 4.5296,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.525,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.5276,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.5342,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.5152,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.5156,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.5121,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736769693709819e-05,
+      "loss": 4.5088,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735931098958767e-05,
+      "loss": 4.5131,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735092504207715e-05,
+      "loss": 4.5168,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734253909456663e-05,
+      "loss": 4.5102,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733415314705611e-05,
+      "loss": 4.5325,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732576719954559e-05,
+      "loss": 4.5123,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.5189,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730901168332828e-05,
+      "loss": 4.5092,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730062573581776e-05,
+      "loss": 4.5184,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729223978830724e-05,
+      "loss": 4.5099,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.5035,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.507,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267098324579415e-05,
+      "loss": 4.5009,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258712377068895e-05,
+      "loss": 4.4857,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250326429558375e-05,
+      "loss": 4.5015,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241940482047855e-05,
+      "loss": 4.5069,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233554534537335e-05,
+      "loss": 4.5024,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722516858702681e-05,
+      "loss": 4.4991,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216799018320024e-05,
+      "loss": 4.5006,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4944,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4899,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719165755459219e-05,
+      "loss": 4.4854,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718327160708167e-05,
+      "loss": 4.4838,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717488565957115e-05,
+      "loss": 4.485,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4847,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4932,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4926,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714135824833281e-05,
+      "loss": 4.4808,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713297230082228e-05,
+      "loss": 4.4881,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4906,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.474,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4876,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4804,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091075320877147e-05,
+      "loss": 4.4622,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082689373366627e-05,
+      "loss": 4.4824,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.4747,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4698,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.4653,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704916196212828e-05,
+      "loss": 4.4685,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4563,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703240644591097e-05,
+      "loss": 4.4859,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702402049840045e-05,
+      "loss": 4.4659,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015634550889924e-05,
+      "loss": 4.4741,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4743,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4579,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4505,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4685,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4539,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965318865826804e-05,
+      "loss": 4.4503,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695693291831629e-05,
+      "loss": 4.4696,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694854697080577e-05,
+      "loss": 4.4663,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694019378090271e-05,
+      "loss": 4.4508,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.4449,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4467,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4565,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4604,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4573,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4598,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4573,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.46,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4458,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.4494,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4526,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683961154718766e-05,
+      "loss": 4.4333,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683122559967714e-05,
+      "loss": 4.4479,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4532,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4513,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4378,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4252,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.4457,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4354,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677254034590723e-05,
+      "loss": 4.4502,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.4401,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675578482968993e-05,
+      "loss": 4.4433,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674739888217941e-05,
+      "loss": 4.437,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673901293466889e-05,
+      "loss": 4.4285,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673062698715837e-05,
+      "loss": 4.4346,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672224103964785e-05,
+      "loss": 4.4323,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.4458,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4479,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.4384,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668869724960576e-05,
+      "loss": 4.4256,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680327680898976e-05,
+      "loss": 4.4291,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.4285,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663572164681665e-05,
+      "loss": 4.435,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655186217171145e-05,
+      "loss": 4.4238,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664680026966063e-05,
+      "loss": 4.4269,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663841432215011e-05,
+      "loss": 4.417,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.436,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.4078,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.4238,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.4166,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.4222,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588115015890714e-05,
+      "loss": 4.416,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579729068380194e-05,
+      "loss": 4.4258,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571343120869674e-05,
+      "loss": 4.4189,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.4147,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655458760465237e-05,
+      "loss": 4.4213,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.4071,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537832088435065e-05,
+      "loss": 4.4257,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652944614092454e-05,
+      "loss": 4.4288,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652106019341402e-05,
+      "loss": 4.414,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65126742459035e-05,
+      "loss": 4.4103,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.4152,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649591872968619e-05,
+      "loss": 4.4153,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648753278217567e-05,
+      "loss": 4.4106,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647914683466515e-05,
+      "loss": 4.408,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.4067,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.4224,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.4115,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.4061,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.3952,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.4075,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.3977,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.4171,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.4126,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3982,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638693416965689e-05,
+      "loss": 4.3967,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637854822214637e-05,
+      "loss": 4.4046,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637016227463585e-05,
+      "loss": 4.4016,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636177632712533e-05,
+      "loss": 4.4127,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.395,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.4049,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.4033,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.4012,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.4005,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.4022,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.3908,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629472150464864e-05,
+      "loss": 4.3963,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628633555713812e-05,
+      "loss": 4.4006,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277949609627597e-05,
+      "loss": 4.3913,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626956366211707e-05,
+      "loss": 4.3889,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626117771460655e-05,
+      "loss": 4.397,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.394,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.364997386932373,
+      "eval_runtime": 291.5075,
+      "eval_samples_per_second": 1309.026,
+      "eval_steps_per_second": 40.908,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244422198389246e-05,
+      "loss": 4.3874,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236036250878726e-05,
+      "loss": 4.3795,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227650303368205e-05,
+      "loss": 4.404,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3843,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.4022,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3758,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.3963,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3803,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.3851,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3879,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3862,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3949,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3797,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3769,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3799,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3659,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.3784,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.382,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.3725,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.3959,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607675238459004e-05,
+      "loss": 4.3805,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606836643707952e-05,
+      "loss": 4.3879,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059980489569e-05,
+      "loss": 4.3762,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605159454205848e-05,
+      "loss": 4.3832,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604320859454796e-05,
+      "loss": 4.378,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603482264703743e-05,
+      "loss": 4.3727,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3758,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601806713082013e-05,
+      "loss": 4.3679,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600968118330961e-05,
+      "loss": 4.3621,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001311614602824e-05,
+      "loss": 4.3665,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5992925667092303e-05,
+      "loss": 4.3766,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984539719581783e-05,
+      "loss": 4.3753,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976153772071257e-05,
+      "loss": 4.3718,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967767824560737e-05,
+      "loss": 4.3745,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3591,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595101230834343e-05,
+      "loss": 4.3635,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942626360832906e-05,
+      "loss": 4.3631,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3539,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3613,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.356,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3678,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590072938088778e-05,
+      "loss": 4.3662,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589234343337726e-05,
+      "loss": 4.3587,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588395748586673e-05,
+      "loss": 4.3581,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587557153835621e-05,
+      "loss": 4.3717,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586718559084569e-05,
+      "loss": 4.3507,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.364,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3565,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.3396,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3632,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582525585329309e-05,
+      "loss": 4.3502,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581686990578257e-05,
+      "loss": 4.3533,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580848395827205e-05,
+      "loss": 4.3423,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800114389565266e-05,
+      "loss": 4.3476,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3377,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.362,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3484,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3541,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3559,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3424,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741429135795355e-05,
+      "loss": 4.3323,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733043188284835e-05,
+      "loss": 4.3522,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572468999838177e-05,
+      "loss": 4.3375,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571630405087125e-05,
+      "loss": 4.3337,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570791810336074e-05,
+      "loss": 4.3513,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569953215585022e-05,
+      "loss": 4.349,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569116258714343e-05,
+      "loss": 4.3373,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568277663963291e-05,
+      "loss": 4.3334,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567439069212239e-05,
+      "loss": 4.3278,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566600474461187e-05,
+      "loss": 4.3383,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565761879710135e-05,
+      "loss": 4.3481,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564923284959083e-05,
+      "loss": 4.3463,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3461,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.342,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624091385863e-05,
+      "loss": 4.3538,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615721817156206e-05,
+      "loss": 4.3301,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560733586964569e-05,
+      "loss": 4.3401,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559894992213517e-05,
+      "loss": 4.3369,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559056397462465e-05,
+      "loss": 4.3199,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558219440591786e-05,
+      "loss": 4.3425,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557380845840734e-05,
+      "loss": 4.341,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556542251089682e-05,
+      "loss": 4.3365,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55570365633863e-05,
+      "loss": 4.3291,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554865061587578e-05,
+      "loss": 4.3183,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554028104716899e-05,
+      "loss": 4.3331,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553189509965847e-05,
+      "loss": 4.3325,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552350915214795e-05,
+      "loss": 4.3383,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551512320463743e-05,
+      "loss": 4.3277,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506770014734376e-05,
+      "loss": 4.3368,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498384067223856e-05,
+      "loss": 4.3248,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489998119713336e-05,
+      "loss": 4.3247,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481612172202816e-05,
+      "loss": 4.3251,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473226224692296e-05,
+      "loss": 4.3225,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464840277181776e-05,
+      "loss": 4.3408,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456454329671256e-05,
+      "loss": 4.3378,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448084760964465e-05,
+      "loss": 4.3376,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439698813453945e-05,
+      "loss": 4.3179,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431312865943425e-05,
+      "loss": 4.3251,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422926918432905e-05,
+      "loss": 4.325,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414540970922385e-05,
+      "loss": 4.3287,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406155023411865e-05,
+      "loss": 4.3202,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397769075901345e-05,
+      "loss": 4.3208,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389383128390824e-05,
+      "loss": 4.3117,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538101355968404e-05,
+      "loss": 4.3324,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372627612173514e-05,
+      "loss": 4.3058,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364241664662994e-05,
+      "loss": 4.3198,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355855717152473e-05,
+      "loss": 4.3164,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347469769641953e-05,
+      "loss": 4.3191,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533908382213143e-05,
+      "loss": 4.3143,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533069787462091e-05,
+      "loss": 4.3265,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532231192711039e-05,
+      "loss": 4.3169,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531392597959987e-05,
+      "loss": 4.3101,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530557278969682e-05,
+      "loss": 4.3242,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52971868421863e-05,
+      "loss": 4.3069,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.3227,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528041494716526e-05,
+      "loss": 4.3295,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527202899965474e-05,
+      "loss": 4.3181,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526364305214422e-05,
+      "loss": 4.311,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525527348343743e-05,
+      "loss": 4.3126,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524688753592691e-05,
+      "loss": 4.3135,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523850158841639e-05,
+      "loss": 4.3132,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523011564090587e-05,
+      "loss": 4.3148,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221746072199076e-05,
+      "loss": 4.3032,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213360124688556e-05,
+      "loss": 4.3227,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5204974177178036e-05,
+      "loss": 4.3159,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196588229667516e-05,
+      "loss": 4.3112,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188202282157e-05,
+      "loss": 4.2951,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517983271345021e-05,
+      "loss": 4.311,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517144676593969e-05,
+      "loss": 4.3021,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516306081842917e-05,
+      "loss": 4.3139,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515467487091865e-05,
+      "loss": 4.3189,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514628892340813e-05,
+      "loss": 4.3071,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513790297589761e-05,
+      "loss": 4.294,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512951702838709e-05,
+      "loss": 4.3091,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512113108087657e-05,
+      "loss": 4.3053,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511277789097351e-05,
+      "loss": 4.3176,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104408322266726e-05,
+      "loss": 4.2971,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096022374756206e-05,
+      "loss": 4.3086,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087636427245685e-05,
+      "loss": 4.3088,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079250479735165e-05,
+      "loss": 4.311,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070864532224645e-05,
+      "loss": 4.3053,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062478584714125e-05,
+      "loss": 4.3061,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054092637203605e-05,
+      "loss": 4.2961,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045706689693085e-05,
+      "loss": 4.3045,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037320742182565e-05,
+      "loss": 4.3046,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028951173475774e-05,
+      "loss": 4.302,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020565225965254e-05,
+      "loss": 4.2936,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012179278454734e-05,
+      "loss": 4.3059,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003793330944214e-05,
+      "loss": 4.2992,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.280091285705566,
+      "eval_runtime": 291.7383,
+      "eval_samples_per_second": 1307.991,
+      "eval_steps_per_second": 40.876,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995407383433694e-05,
+      "loss": 4.2914,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987021435923174e-05,
+      "loss": 4.2903,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978635488412654e-05,
+      "loss": 4.3098,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970249540902134e-05,
+      "loss": 4.2949,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961863593391614e-05,
+      "loss": 4.3056,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953477645881094e-05,
+      "loss": 4.293,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945091698370574e-05,
+      "loss": 4.2998,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936705750860054e-05,
+      "loss": 4.2937,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928319803349534e-05,
+      "loss": 4.2937,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919933855839014e-05,
+      "loss": 4.2961,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911547908328494e-05,
+      "loss": 4.2971,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903161960817974e-05,
+      "loss": 4.3047,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489479239211118e-05,
+      "loss": 4.2893,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488642282340439e-05,
+      "loss": 4.2885,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487803687589387e-05,
+      "loss": 4.2894,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486965092838336e-05,
+      "loss": 4.2791,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486126498087284e-05,
+      "loss": 4.2889,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485287903336232e-05,
+      "loss": 4.2912,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48444930858518e-05,
+      "loss": 4.2863,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483610713834128e-05,
+      "loss": 4.3072,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482772119083076e-05,
+      "loss": 4.2942,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481933524332024e-05,
+      "loss": 4.3011,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481094929580971e-05,
+      "loss": 4.2907,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480256334829919e-05,
+      "loss": 4.2957,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479417740078867e-05,
+      "loss": 4.2894,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478579145327815e-05,
+      "loss": 4.2911,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477740550576763e-05,
+      "loss": 4.2876,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476903593706084e-05,
+      "loss": 4.2826,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476064998955033e-05,
+      "loss": 4.2799,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752280420843536e-05,
+      "loss": 4.2802,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743894473333016e-05,
+      "loss": 4.2878,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735508525822496e-05,
+      "loss": 4.2911,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727122578311976e-05,
+      "loss": 4.2863,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718736630801456e-05,
+      "loss": 4.288,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710367062094665e-05,
+      "loss": 4.2758,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701981114584145e-05,
+      "loss": 4.2807,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693595167073625e-05,
+      "loss": 4.2808,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685209219563105e-05,
+      "loss": 4.267,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676823272052585e-05,
+      "loss": 4.28,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668437324542065e-05,
+      "loss": 4.271,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466006775583528e-05,
+      "loss": 4.2857,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465169818712849e-05,
+      "loss": 4.2845,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464331223961797e-05,
+      "loss": 4.2792,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463492629210745e-05,
+      "loss": 4.2729,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462654034459693e-05,
+      "loss": 4.2897,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461815439708641e-05,
+      "loss": 4.2672,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460976844957589e-05,
+      "loss": 4.2828,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460138250206537e-05,
+      "loss": 4.273,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459299655455485e-05,
+      "loss": 4.2579,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458461060704433e-05,
+      "loss": 4.2838,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457624103833754e-05,
+      "loss": 4.2696,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456787146963075e-05,
+      "loss": 4.2699,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559485522120234e-05,
+      "loss": 4.2647,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551099574609714e-05,
+      "loss": 4.2618,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542713627099194e-05,
+      "loss": 4.2626,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534327679588674e-05,
+      "loss": 4.2736,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525941732078154e-05,
+      "loss": 4.2718,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517555784567634e-05,
+      "loss": 4.2741,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509169837057114e-05,
+      "loss": 4.2763,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450080026835032e-05,
+      "loss": 4.2652,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44924143208398e-05,
+      "loss": 4.2525,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448402837332928e-05,
+      "loss": 4.2748,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447564242581876e-05,
+      "loss": 4.2584,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446725647830824e-05,
+      "loss": 4.2535,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458870530797716e-05,
+      "loss": 4.2751,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44504845832872e-05,
+      "loss": 4.2692,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444209863577668e-05,
+      "loss": 4.2597,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443371268826616e-05,
+      "loss": 4.2583,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442534311955937e-05,
+      "loss": 4.2466,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441695717204885e-05,
+      "loss": 4.26,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440857122453833e-05,
+      "loss": 4.2733,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440018527702781e-05,
+      "loss": 4.268,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439179932951729e-05,
+      "loss": 4.2668,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43834297608105e-05,
+      "loss": 4.2633,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437504381329998e-05,
+      "loss": 4.281,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436667424459319e-05,
+      "loss": 4.2543,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2656,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2649,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2387,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2715,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.264,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316358559530076e-05,
+      "loss": 4.2631,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307972612019556e-05,
+      "loss": 4.2568,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299603043312766e-05,
+      "loss": 4.2401,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291217095802245e-05,
+      "loss": 4.2586,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282831148291725e-05,
+      "loss": 4.2567,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274445200781205e-05,
+      "loss": 4.2654,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266059253270685e-05,
+      "loss": 4.2533,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257706063367624e-05,
+      "loss": 4.2624,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249320115857104e-05,
+      "loss": 4.25,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424093416834659e-05,
+      "loss": 4.2542,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423254822083607e-05,
+      "loss": 4.249,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422416227332555e-05,
+      "loss": 4.2492,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421577632581503e-05,
+      "loss": 4.2661,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420739037830451e-05,
+      "loss": 4.2631,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419900443079399e-05,
+      "loss": 4.2683,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41906348620872e-05,
+      "loss": 4.2423,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418224891457668e-05,
+      "loss": 4.2533,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417386296706616e-05,
+      "loss": 4.2532,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416547701955564e-05,
+      "loss": 4.2546,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415709107204512e-05,
+      "loss": 4.2514,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414872150333833e-05,
+      "loss": 4.2441,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414033555582781e-05,
+      "loss": 4.2452,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413194960831729e-05,
+      "loss": 4.2605,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123580039610504e-05,
+      "loss": 4.2331,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115194092099984e-05,
+      "loss": 4.2459,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410682452339319e-05,
+      "loss": 4.2399,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409843857588267e-05,
+      "loss": 4.2535,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409005262837215e-05,
+      "loss": 4.2426,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408166668086163e-05,
+      "loss": 4.2536,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407328073335111e-05,
+      "loss": 4.2455,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406489478584059e-05,
+      "loss": 4.2395,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405650883833007e-05,
+      "loss": 4.2529,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404812289081955e-05,
+      "loss": 4.239,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039736943309026e-05,
+      "loss": 4.2535,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403135099579851e-05,
+      "loss": 4.2569,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402296504828799e-05,
+      "loss": 4.2456,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401457910077747e-05,
+      "loss": 4.2398,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400619315326695e-05,
+      "loss": 4.2449,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399780720575643e-05,
+      "loss": 4.2412,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.2453,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.245,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39726657420286e-05,
+      "loss": 4.2356,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396427979451808e-05,
+      "loss": 4.2529,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.2452,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.2468,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393913833079025e-05,
+      "loss": 4.2273,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393075238327973e-05,
+      "loss": 4.2415,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922382814572946e-05,
+      "loss": 4.2332,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913996867062426e-05,
+      "loss": 4.2446,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905610919551906e-05,
+      "loss": 4.2477,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.2418,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888855403334595e-05,
+      "loss": 4.2252,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880469455824075e-05,
+      "loss": 4.2442,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872083508313555e-05,
+      "loss": 4.2376,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863697560803035e-05,
+      "loss": 4.2503,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855311613292515e-05,
+      "loss": 4.2293,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846942044585724e-05,
+      "loss": 4.2453,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838556097075204e-05,
+      "loss": 4.2414,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830170149564684e-05,
+      "loss": 4.2452,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821784202054164e-05,
+      "loss": 4.236,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813398254543644e-05,
+      "loss": 4.2395,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380501230703313e-05,
+      "loss": 4.2281,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2398,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2374,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37798708433053e-05,
+      "loss": 4.2346,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377150127459851e-05,
+      "loss": 4.2285,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376311532708799e-05,
+      "loss": 4.2372,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375472937957747e-05,
+      "loss": 4.2379,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.222952365875244,
+      "eval_runtime": 293.7245,
+      "eval_samples_per_second": 1299.146,
+      "eval_steps_per_second": 40.599,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374634343206695e-05,
+      "loss": 4.2327,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373795748455643e-05,
+      "loss": 4.2243,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372957153704591e-05,
+      "loss": 4.2436,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372120196833912e-05,
+      "loss": 4.2291,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3712832399632333e-05,
+      "loss": 4.2384,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704446452121813e-05,
+      "loss": 4.2301,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369606050461129e-05,
+      "loss": 4.2308,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368767455710077e-05,
+      "loss": 4.2321,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367928860959025e-05,
+      "loss": 4.2272,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367091904088346e-05,
+      "loss": 4.2331,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366253309337294e-05,
+      "loss": 4.2384,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365414714586242e-05,
+      "loss": 4.236,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364577757715563e-05,
+      "loss": 4.2278,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363739162964511e-05,
+      "loss": 4.2243,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362900568213459e-05,
+      "loss": 4.2282,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362061973462407e-05,
+      "loss": 4.2135,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361223378711355e-05,
+      "loss": 4.2254,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360384783960304e-05,
+      "loss": 4.2254,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359546189209251e-05,
+      "loss": 4.2274,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358707594458199e-05,
+      "loss": 4.2438,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357868999707147e-05,
+      "loss": 4.2336,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357030404956095e-05,
+      "loss": 4.2376,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356191810205043e-05,
+      "loss": 4.2296,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355353215453991e-05,
+      "loss": 4.2308,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354516258583312e-05,
+      "loss": 4.2281,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35367766383226e-05,
+      "loss": 4.2287,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352839069081208e-05,
+      "loss": 4.2227,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352000474330156e-05,
+      "loss": 4.2218,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351161879579104e-05,
+      "loss": 4.2191,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350323284828052e-05,
+      "loss": 4.2194,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494863279573736e-05,
+      "loss": 4.2243,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486477332063216e-05,
+      "loss": 4.2279,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478091384552696e-05,
+      "loss": 4.2245,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469705437042175e-05,
+      "loss": 4.2305,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461319489531655e-05,
+      "loss": 4.2134,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452949920824865e-05,
+      "loss": 4.2216,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444563973314344e-05,
+      "loss": 4.2177,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436178025803824e-05,
+      "loss": 4.2085,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427792078293304e-05,
+      "loss": 4.2211,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419422509586514e-05,
+      "loss": 4.204,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411036562075993e-05,
+      "loss": 4.2227,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402650614565473e-05,
+      "loss": 4.2217,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339426466705496e-05,
+      "loss": 4.2231,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338587871954444e-05,
+      "loss": 4.2142,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337749277203392e-05,
+      "loss": 4.2301,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33691068245234e-05,
+      "loss": 4.2063,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336073725581661e-05,
+      "loss": 4.2243,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335235130830609e-05,
+      "loss": 4.2133,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334396536079557e-05,
+      "loss": 4.1928,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333557941328505e-05,
+      "loss": 4.2276,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332720984457826e-05,
+      "loss": 4.209,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331882389706774e-05,
+      "loss": 4.216,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331043794955722e-05,
+      "loss": 4.2031,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330206838085043e-05,
+      "loss": 4.2026,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293682433339914e-05,
+      "loss": 4.2087,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285296485829394e-05,
+      "loss": 4.2103,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276910538318874e-05,
+      "loss": 4.2154,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326852459080835e-05,
+      "loss": 4.2125,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326013864329783e-05,
+      "loss": 4.2234,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325175269578731e-05,
+      "loss": 4.2058,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324336674827679e-05,
+      "loss": 4.1964,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323498080076627e-05,
+      "loss": 4.2151,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322659485325575e-05,
+      "loss": 4.1958,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321820890574523e-05,
+      "loss": 4.1977,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209822958234707e-05,
+      "loss": 4.2164,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201453389527916e-05,
+      "loss": 4.2141,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193067442017396e-05,
+      "loss": 4.1997,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184681494506876e-05,
+      "loss": 4.2032,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317629554699636e-05,
+      "loss": 4.192,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316792597828957e-05,
+      "loss": 4.2023,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315955640958278e-05,
+      "loss": 4.2145,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315117046207226e-05,
+      "loss": 4.2104,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314278451456174e-05,
+      "loss": 4.2096,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313439856705122e-05,
+      "loss": 4.2076,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31260126195407e-05,
+      "loss": 4.2227,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311762667203018e-05,
+      "loss": 4.1957,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310924072451966e-05,
+      "loss": 4.2094,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310087115581287e-05,
+      "loss": 4.2064,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309248520830235e-05,
+      "loss": 4.184,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308409926079183e-05,
+      "loss": 4.2126,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075729692085045e-05,
+      "loss": 4.2092,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067343744574525e-05,
+      "loss": 4.2061,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058957797064005e-05,
+      "loss": 4.204,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050571849553485e-05,
+      "loss": 4.1892,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042185902042965e-05,
+      "loss": 4.1957,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033799954532445e-05,
+      "loss": 4.2018,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025430385825654e-05,
+      "loss": 4.2091,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017044438315134e-05,
+      "loss": 4.1959,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008658490804614e-05,
+      "loss": 4.2082,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000272543294094e-05,
+      "loss": 4.1948,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991886595783574e-05,
+      "loss": 4.1983,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983500648273054e-05,
+      "loss": 4.1928,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975114700762534e-05,
+      "loss": 4.1959,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296674513205575e-05,
+      "loss": 4.2108,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295835918454523e-05,
+      "loss": 4.2094,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294997323703471e-05,
+      "loss": 4.2134,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294158728952418e-05,
+      "loss": 4.1864,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293320134201366e-05,
+      "loss": 4.2002,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292481539450314e-05,
+      "loss": 4.1941,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291642944699262e-05,
+      "loss": 4.2047,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29080434994821e-05,
+      "loss": 4.1956,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289965755197158e-05,
+      "loss": 4.1949,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289127160446106e-05,
+      "loss": 4.1878,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288288565695054e-05,
+      "loss": 4.2035,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287449970944002e-05,
+      "loss": 4.1817,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286613014073324e-05,
+      "loss": 4.198,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285776057202645e-05,
+      "loss": 4.181,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284937462451593e-05,
+      "loss": 4.2019,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284098867700541e-05,
+      "loss": 4.1924,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283260272949489e-05,
+      "loss": 4.1943,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282421678198437e-05,
+      "loss": 4.1921,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281583083447385e-05,
+      "loss": 4.1936,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280744488696333e-05,
+      "loss": 4.1951,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279905893945281e-05,
+      "loss": 4.1896,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790689370746016e-05,
+      "loss": 4.1962,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782303423235496e-05,
+      "loss": 4.2088,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773933854528705e-05,
+      "loss": 4.1933,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276554790701819e-05,
+      "loss": 4.1884,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275716195950767e-05,
+      "loss": 4.1939,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274877601199715e-05,
+      "loss": 4.193,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274039006448663e-05,
+      "loss": 4.1871,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273202049577984e-05,
+      "loss": 4.1953,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272363454826932e-05,
+      "loss": 4.182,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27152486007588e-05,
+      "loss": 4.1984,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270686265324828e-05,
+      "loss": 4.1975,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269847670573776e-05,
+      "loss": 4.1902,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269009075822724e-05,
+      "loss": 4.1805,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681704810716714e-05,
+      "loss": 4.1869,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673318863206194e-05,
+      "loss": 4.1863,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266494929449941e-05,
+      "loss": 4.1958,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265656334698889e-05,
+      "loss": 4.1912,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264817739947837e-05,
+      "loss": 4.1934,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263979145196785e-05,
+      "loss": 4.1721,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631421883261065e-05,
+      "loss": 4.194,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623052314554274e-05,
+      "loss": 4.1872,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614666367043754e-05,
+      "loss": 4.1966,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606280419533234e-05,
+      "loss": 4.1793,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597894472022714e-05,
+      "loss": 4.1986,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.1882,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.1943,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.182,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1965,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255598111327384e-05,
+      "loss": 4.1738,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254759516576332e-05,
+      "loss": 4.189,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253922559705654e-05,
+      "loss": 4.1896,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253083964954602e-05,
+      "loss": 4.1809,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252245370203549e-05,
+      "loss": 4.1844,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251408413332871e-05,
+      "loss": 4.1808,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250569818581819e-05,
+      "loss": 4.1886,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.180971145629883,
+      "eval_runtime": 292.488,
+      "eval_samples_per_second": 1304.638,
+      "eval_steps_per_second": 40.771,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249731223830766e-05,
+      "loss": 4.1858,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248892629079714e-05,
+      "loss": 4.174,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248054034328662e-05,
+      "loss": 4.1913,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24721543957761e-05,
+      "loss": 4.1791,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246376844826558e-05,
+      "loss": 4.1895,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245538250075506e-05,
+      "loss": 4.1804,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244699655324455e-05,
+      "loss": 4.1819,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243861060573403e-05,
+      "loss": 4.1859,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243022465822351e-05,
+      "loss": 4.1805,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242185508951672e-05,
+      "loss": 4.1836,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24134691420062e-05,
+      "loss": 4.1891,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240508319449568e-05,
+      "loss": 4.1868,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239669724698516e-05,
+      "loss": 4.1802,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388311299474637e-05,
+      "loss": 4.1758,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379925351964117e-05,
+      "loss": 4.18,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371539404453596e-05,
+      "loss": 4.1618,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363153456943076e-05,
+      "loss": 4.1797,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235476750943255e-05,
+      "loss": 4.1776,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234638156192203e-05,
+      "loss": 4.1762,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337995614411516e-05,
+      "loss": 4.1964,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329609666900996e-05,
+      "loss": 4.1876,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.1922,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312854150683685e-05,
+      "loss": 4.1775,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304468203173165e-05,
+      "loss": 4.1829,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296098634466374e-05,
+      "loss": 4.1804,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287712686955854e-05,
+      "loss": 4.1792,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279326739445334e-05,
+      "loss": 4.1775,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270940791934814e-05,
+      "loss": 4.177,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262554844424294e-05,
+      "loss": 4.1688,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254168896913774e-05,
+      "loss": 4.1712,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245782949403254e-05,
+      "loss": 4.1752,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237397001892734e-05,
+      "loss": 4.1841,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222902743318595e-05,
+      "loss": 4.1774,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222064148567543e-05,
+      "loss": 4.1841,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221225553816491e-05,
+      "loss": 4.1684,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220386959065439e-05,
+      "loss": 4.1708,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219548364314387e-05,
+      "loss": 4.1736,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218711407443708e-05,
+      "loss": 4.162,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217872812692656e-05,
+      "loss": 4.1711,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217034217941604e-05,
+      "loss": 4.1651,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216195623190552e-05,
+      "loss": 4.173,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153570284395e-05,
+      "loss": 4.1763,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214520071568821e-05,
+      "loss": 4.1731,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213681476817769e-05,
+      "loss": 4.1693,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212842882066717e-05,
+      "loss": 4.1808,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212004287315665e-05,
+      "loss": 4.1652,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111656925646134e-05,
+      "loss": 4.1772,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1676,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094901409428823e-05,
+      "loss": 4.1451,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20865154619183e-05,
+      "loss": 4.1889,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207812951440778e-05,
+      "loss": 4.1605,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206975994570099e-05,
+      "loss": 4.172,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206137399819047e-05,
+      "loss": 4.1536,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205298805067995e-05,
+      "loss": 4.161,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204460210316943e-05,
+      "loss": 4.1571,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203621615565891e-05,
+      "loss": 4.1677,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027830208148385e-05,
+      "loss": 4.1677,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201944426063787e-05,
+      "loss": 4.1696,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201105831312735e-05,
+      "loss": 4.1767,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200267236561683e-05,
+      "loss": 4.1637,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199430279691004e-05,
+      "loss": 4.1503,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198591684939952e-05,
+      "loss": 4.1755,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977530901889e-05,
+      "loss": 4.1509,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196914495437848e-05,
+      "loss": 4.1485,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196075900686796e-05,
+      "loss": 4.1718,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195238943816117e-05,
+      "loss": 4.1714,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194400349065065e-05,
+      "loss": 4.1544,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193561754314013e-05,
+      "loss": 4.1622,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192723159562961e-05,
+      "loss": 4.1474,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191884564811909e-05,
+      "loss": 4.1587,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910476079412306e-05,
+      "loss": 4.1672,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.166,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.1676,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1659,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876932289370226e-05,
+      "loss": 4.1747,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868546341859706e-05,
+      "loss": 4.1508,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860160394349186e-05,
+      "loss": 4.1682,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851790825642395e-05,
+      "loss": 4.1591,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843404878131875e-05,
+      "loss": 4.1483,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835018930621355e-05,
+      "loss": 4.1637,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826632983110834e-05,
+      "loss": 4.1672,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818263414404044e-05,
+      "loss": 4.1639,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809877466893524e-05,
+      "loss": 4.1612,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180149151938301e-05,
+      "loss": 4.146,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179310557187249e-05,
+      "loss": 4.1503,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178471962436197e-05,
+      "loss": 4.1616,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177633367685145e-05,
+      "loss": 4.1653,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1536,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1641,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175119221312362e-05,
+      "loss": 4.1448,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17428062656131e-05,
+      "loss": 4.1597,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173443669690631e-05,
+      "loss": 4.1527,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172605074939579e-05,
+      "loss": 4.155,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171766480188527e-05,
+      "loss": 4.1684,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170927885437475e-05,
+      "loss": 4.1681,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170089290686423e-05,
+      "loss": 4.1689,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169250695935371e-05,
+      "loss": 4.1448,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684137390646924e-05,
+      "loss": 4.1592,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16757514431364e-05,
+      "loss": 4.1496,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166736549562588e-05,
+      "loss": 4.1634,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165897954811536e-05,
+      "loss": 4.1534,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165059360060484e-05,
+      "loss": 4.1559,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164220765309432e-05,
+      "loss": 4.1474,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1577,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625452136877006e-05,
+      "loss": 4.1435,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617066189366486e-05,
+      "loss": 4.1554,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608680241855966e-05,
+      "loss": 4.1398,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600294294345446e-05,
+      "loss": 4.1615,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159192472563866e-05,
+      "loss": 4.1497,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1525,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157515283061762e-05,
+      "loss": 4.1489,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.1553,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155838093559658e-05,
+      "loss": 4.1541,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154999498808606e-05,
+      "loss": 4.1485,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154160904057554e-05,
+      "loss": 4.1572,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153322309306502e-05,
+      "loss": 4.1645,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152486990316196e-05,
+      "loss": 4.157,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151648395565144e-05,
+      "loss": 4.1499,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150809800814092e-05,
+      "loss": 4.1469,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14997120606304e-05,
+      "loss": 4.1542,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149132611311988e-05,
+      "loss": 4.1448,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482956544413095e-05,
+      "loss": 4.1568,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474570596902575e-05,
+      "loss": 4.1396,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.1591,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457815080685264e-05,
+      "loss": 4.1553,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449429133174744e-05,
+      "loss": 4.1515,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441043185664224e-05,
+      "loss": 4.1408,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1461,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424271290643184e-05,
+      "loss": 4.1446,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.1582,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407499395622144e-05,
+      "loss": 4.1518,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399113448111624e-05,
+      "loss": 4.15,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139074387940483e-05,
+      "loss": 4.133,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138237431069805e-05,
+      "loss": 4.1547,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137398836318753e-05,
+      "loss": 4.1464,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.1597,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1414,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1561,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134046095194918e-05,
+      "loss": 4.1468,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133207500443866e-05,
+      "loss": 4.1616,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132370543573187e-05,
+      "loss": 4.1411,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131531948822135e-05,
+      "loss": 4.1523,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130693354071083e-05,
+      "loss": 4.1367,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129854759320031e-05,
+      "loss": 4.1492,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129016164568979e-05,
+      "loss": 4.1493,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281775698179273e-05,
+      "loss": 4.14,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273389750668753e-05,
+      "loss": 4.1501,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126500380315823e-05,
+      "loss": 4.139,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125663423445144e-05,
+      "loss": 4.153,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.150570869445801,
+      "eval_runtime": 297.8513,
+      "eval_samples_per_second": 1281.146,
+      "eval_steps_per_second": 40.037,
+      "step": 534240
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.19517830415354e+17,
+  "trial_name": null,
+  "trial_params": null
+}