diff --git "a/checkpoint-1068480/trainer_state.json" "b/checkpoint-1068480/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1068480/trainer_state.json"
@@ -0,0 +1,12653 @@
+{
+  "best_metric": 3.861919403076172,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/transformer/2/checkpoints/checkpoint-839520",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1068480,
+  "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": 10.9339,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.7961,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1583,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9557,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.7951,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.696,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5911,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5362,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4416,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4026,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3441,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3218,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2749,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2218,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1892,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1384,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1165,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0927,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0637,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0354,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0224,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9923,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9717,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9355,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9347,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.9029,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8898,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8709,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.845,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8283,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8102,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.8038,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7855,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7808,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7658,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7542,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7417,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7058,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7041,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.693,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6758,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6793,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 4.6739,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.6516,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6505,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6436,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6353,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6317,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6123,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 4.5983,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 4.6005,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 4.5863,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.587,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5597,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5737,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5478,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5683,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.543,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5395,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 4.5312,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 4.5263,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 4.5128,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 4.5184,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 4.4938,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 4.4876,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5015,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4921,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 4.4802,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 4.4651,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 4.4673,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 4.4738,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 4.468,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 4.4601,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 4.4587,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 4.46,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 4.4524,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 4.4301,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 4.4314,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 4.4306,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 4.4217,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 4.4218,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 4.4141,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 4.4306,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4069,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 4.3908,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 4.4015,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 4.3978,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 4.4028,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 4.3857,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 4.3858,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 4.384,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 4.3776,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 4.3665,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 4.3688,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 4.3716,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 4.3789,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.3576,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.3496,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.3499,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.3495,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.3549,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.3403,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.3343,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.3277,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.3382,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.3062,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.3264,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.3179,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.3246,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.3134,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.3177,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.3067,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.3072,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.3117,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.2927,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.3113,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.3076,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.3007,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.2902,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.2891,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.2905,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.2878,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.287,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.2768,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.2846,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.2669,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.2704,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.2597,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.2633,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.2529,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.2745,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.2769,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.2628,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.2435,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766433672724e-05,
+      "loss": 4.2621,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886825741976188e-05,
+      "loss": 4.2561,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.253,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.2448,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.2504,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.2454,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.241,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.2422,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.2441,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.2301,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.2319,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.2429,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.2291,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.2285,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.2395,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.2207,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2148919105529785,
+      "eval_runtime": 304.5041,
+      "eval_samples_per_second": 1253.155,
+      "eval_steps_per_second": 39.162,
+      "step": 76320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.2011,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.212,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872574544849423e-05,
+      "loss": 4.2259,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871735950098371e-05,
+      "loss": 4.2144,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870897355347319e-05,
+      "loss": 4.2212,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870058760596267e-05,
+      "loss": 4.2,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869220165845215e-05,
+      "loss": 4.2117,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868381571094163e-05,
+      "loss": 4.1911,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867542976343111e-05,
+      "loss": 4.2081,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866704381592059e-05,
+      "loss": 4.192,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865865786841007e-05,
+      "loss": 4.2064,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865027192089955e-05,
+      "loss": 4.2017,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864190235219277e-05,
+      "loss": 4.1937,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863351640468225e-05,
+      "loss": 4.1842,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862513045717173e-05,
+      "loss": 4.1761,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861674450966121e-05,
+      "loss": 4.1766,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8608374940954416e-05,
+      "loss": 4.1844,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8599988993443896e-05,
+      "loss": 4.1834,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8591603045933376e-05,
+      "loss": 4.1753,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583217098422856e-05,
+      "loss": 4.1977,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574847529716065e-05,
+      "loss": 4.1866,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566461582205545e-05,
+      "loss": 4.1784,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558075634695025e-05,
+      "loss": 4.1716,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8549689687184505e-05,
+      "loss": 4.1891,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541303739673985e-05,
+      "loss": 4.162,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532917792163465e-05,
+      "loss": 4.175,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524531844652945e-05,
+      "loss": 4.1718,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516145897142425e-05,
+      "loss": 4.1649,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850777632843564e-05,
+      "loss": 4.1566,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499390380925114e-05,
+      "loss": 4.1548,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849102081221833e-05,
+      "loss": 4.1611,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848263486470781e-05,
+      "loss": 4.1593,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847424891719729e-05,
+      "loss": 4.1606,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846586296968676e-05,
+      "loss": 4.1602,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845747702217624e-05,
+      "loss": 4.1594,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844909107466572e-05,
+      "loss": 4.1543,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844072150595894e-05,
+      "loss": 4.1362,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843233555844842e-05,
+      "loss": 4.1444,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84239496109379e-05,
+      "loss": 4.138,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841556366342738e-05,
+      "loss": 4.1334,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.1436,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839880814721007e-05,
+      "loss": 4.1456,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839042219969955e-05,
+      "loss": 4.1335,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838203625218903e-05,
+      "loss": 4.1326,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8373666683482236e-05,
+      "loss": 4.1365,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8365280735971716e-05,
+      "loss": 4.1362,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8356894788461196e-05,
+      "loss": 4.1386,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8348508840950676e-05,
+      "loss": 4.1276,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.1241,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833175332473337e-05,
+      "loss": 4.1293,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832336737722285e-05,
+      "loss": 4.1215,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.831498142971233e-05,
+      "loss": 4.127,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830659548220181e-05,
+      "loss": 4.1133,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829822591349502e-05,
+      "loss": 4.1268,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.82898399659845e-05,
+      "loss": 4.1025,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828145401847398e-05,
+      "loss": 4.1272,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.827306807096346e-05,
+      "loss": 4.1137,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.1114,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.1201,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.1086,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.1014,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.1158,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.0902,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.0912,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.1074,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8197643679779975e-05,
+      "loss": 4.1053,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8189257732269455e-05,
+      "loss": 4.0985,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8180871784758935e-05,
+      "loss": 4.0878,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8172485837248415e-05,
+      "loss": 4.0884,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.0957,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.1038,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.0941,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.0994,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.1002,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.098,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.08,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.0841,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.091,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.0803,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.0854,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807190360353337e-05,
+      "loss": 4.0773,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806351765602285e-05,
+      "loss": 4.0987,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.805513170851233e-05,
+      "loss": 4.0801,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804676213980554e-05,
+      "loss": 4.0679,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803837619229502e-05,
+      "loss": 4.0769,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8029990244784504e-05,
+      "loss": 4.0793,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021604297273984e-05,
+      "loss": 4.085,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.0706,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.0712,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799646283354615e-05,
+      "loss": 4.0796,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798807688603563e-05,
+      "loss": 4.0681,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.0632,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.0679,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.074,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.0774,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.0706,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.0543,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.0609,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.0635,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.0654,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.0621,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.0562,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.0488,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.0654,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.0299,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.0545,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.0469,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.0592,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.044,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7828809398550675e-05,
+      "loss": 4.0537,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.0421,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.0446,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.0539,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.0393,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.0541,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.0553,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.0464,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.0398,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.0433,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.0452,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.0339,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.0476,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.0323,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.043,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.0338,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.0305,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.0214,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.0268,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.0191,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.042,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.0496,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.0332,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7636030878631096e-05,
+      "loss": 4.0169,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.0364,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.0267,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.0298,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760250346739275e-05,
+      "loss": 4.0253,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759411751988223e-05,
+      "loss": 4.0276,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.0225,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.0272,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75689760561544e-05,
+      "loss": 4.0232,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756059010864388e-05,
+      "loss": 4.0284,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755220416113336e-05,
+      "loss": 4.0222,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754381821362284e-05,
+      "loss": 4.0141,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753544864491605e-05,
+      "loss": 4.0302,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7527062697405536e-05,
+      "loss": 4.0231,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7518676749895016e-05,
+      "loss": 4.0147,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7510290802384496e-05,
+      "loss": 4.0301,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.0118,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.044399738311768,
+      "eval_runtime": 302.1723,
+      "eval_samples_per_second": 1262.826,
+      "eval_steps_per_second": 39.464,
+      "step": 152640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7493535286167185e-05,
+      "loss": 4.003,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7485149338656665e-05,
+      "loss": 4.0051,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7476763391146145e-05,
+      "loss": 4.022,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7468393822439354e-05,
+      "loss": 4.0145,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7460007874928834e-05,
+      "loss": 4.0197,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7451621927418314e-05,
+      "loss": 4.001,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7443235979907794e-05,
+      "loss": 4.0149,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 3.9915,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.742648046369049e-05,
+      "loss": 4.0149,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.741809451617997e-05,
+      "loss": 3.9956,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740970856866945e-05,
+      "loss": 4.0134,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.0085,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.739296943125587e-05,
+      "loss": 4.0023,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.738458348374535e-05,
+      "loss": 3.9932,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.737619753623483e-05,
+      "loss": 3.9905,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.736781158872431e-05,
+      "loss": 3.9914,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735942564121379e-05,
+      "loss": 3.9955,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735103969370327e-05,
+      "loss": 3.9989,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.734265374619275e-05,
+      "loss": 3.9951,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733426779868223e-05,
+      "loss": 4.011,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7325898229975444e-05,
+      "loss": 4.0051,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7317512282464924e-05,
+      "loss": 4.0008,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7309126334954404e-05,
+      "loss": 3.9946,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7300740387443884e-05,
+      "loss": 4.0035,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.729237081873709e-05,
+      "loss": 3.9872,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.728398487122657e-05,
+      "loss": 3.9961,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.727559892371605e-05,
+      "loss": 3.9947,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.726722935500926e-05,
+      "loss": 3.9909,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.725884340749874e-05,
+      "loss": 3.9795,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.725045745998822e-05,
+      "loss": 3.9834,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.72420715124777e-05,
+      "loss": 3.9862,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.723370194377091e-05,
+      "loss": 3.9909,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.72253159962604e-05,
+      "loss": 3.994,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 3.9866,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720854410123936e-05,
+      "loss": 3.9871,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720015815372884e-05,
+      "loss": 3.9854,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719177220621832e-05,
+      "loss": 3.9711,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71833862587078e-05,
+      "loss": 3.9805,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7175016690001006e-05,
+      "loss": 3.9702,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9703,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.976,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9855,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7141489278762655e-05,
+      "loss": 3.9736,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.9691,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7124717383741615e-05,
+      "loss": 3.9826,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.9718,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.710796186752431e-05,
+      "loss": 3.9828,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9664,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.709118997250327e-05,
+      "loss": 3.9637,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.708282040379648e-05,
+      "loss": 3.9772,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.707443445628596e-05,
+      "loss": 3.9674,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.706604850877544e-05,
+      "loss": 3.9725,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.705766256126492e-05,
+      "loss": 3.9607,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704927661375439e-05,
+      "loss": 3.9682,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704090704504761e-05,
+      "loss": 3.9538,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.703252109753709e-05,
+      "loss": 3.9687,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.702413515002657e-05,
+      "loss": 3.9623,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.701574920251605e-05,
+      "loss": 3.9639,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.700736325500553e-05,
+      "loss": 3.9711,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.699897730749501e-05,
+      "loss": 3.9553,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.699059135998449e-05,
+      "loss": 3.9521,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.698220541247397e-05,
+      "loss": 3.9717,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.697383584376718e-05,
+      "loss": 3.9432,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.696544989625666e-05,
+      "loss": 3.9437,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.695706394874614e-05,
+      "loss": 3.9592,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694867800123562e-05,
+      "loss": 3.9651,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694030843252883e-05,
+      "loss": 3.9553,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6931922485018307e-05,
+      "loss": 3.9435,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6923536537507787e-05,
+      "loss": 3.9471,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6915166968801e-05,
+      "loss": 3.9482,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.690678102129048e-05,
+      "loss": 3.9621,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.689839507377996e-05,
+      "loss": 3.9537,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.689000912626944e-05,
+      "loss": 3.957,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688163955756265e-05,
+      "loss": 3.9571,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.687325361005213e-05,
+      "loss": 3.9608,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.686488404134534e-05,
+      "loss": 3.9424,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.685649809383482e-05,
+      "loss": 3.9443,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.68481121463243e-05,
+      "loss": 3.9544,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.683972619881378e-05,
+      "loss": 3.9399,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.683134025130326e-05,
+      "loss": 3.95,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.682295430379274e-05,
+      "loss": 3.9412,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.681456835628222e-05,
+      "loss": 3.9586,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.680618240877171e-05,
+      "loss": 3.9457,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6797812840064916e-05,
+      "loss": 3.9344,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6789426892554396e-05,
+      "loss": 3.9392,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6781040945043876e-05,
+      "loss": 3.9451,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6772654997533356e-05,
+      "loss": 3.9552,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6764285428826565e-05,
+      "loss": 3.9361,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6755899481316045e-05,
+      "loss": 3.94,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6747513533805525e-05,
+      "loss": 3.9379,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6739127586295005e-05,
+      "loss": 3.9457,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6730758017588214e-05,
+      "loss": 3.9322,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6722372070077694e-05,
+      "loss": 3.9339,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6713986122567174e-05,
+      "loss": 3.9446,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.670561655386039e-05,
+      "loss": 3.9474,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669723060634987e-05,
+      "loss": 3.9439,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668884465883935e-05,
+      "loss": 3.9279,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668045871132883e-05,
+      "loss": 3.9318,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.667207276381831e-05,
+      "loss": 3.9324,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.666368681630779e-05,
+      "loss": 3.938,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.665530086879727e-05,
+      "loss": 3.938,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.664691492128675e-05,
+      "loss": 3.9294,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663854535257996e-05,
+      "loss": 3.9246,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9388,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9108,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9271,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6605017941341614e-05,
+      "loss": 3.9219,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6596631993831094e-05,
+      "loss": 3.9357,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.65882624251243e-05,
+      "loss": 3.9246,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.657987647761378e-05,
+      "loss": 3.929,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.657149053010326e-05,
+      "loss": 3.9196,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.656310458259274e-05,
+      "loss": 3.9239,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.655471863508222e-05,
+      "loss": 3.9323,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.65463326875717e-05,
+      "loss": 3.9226,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6537946740061176e-05,
+      "loss": 3.9322,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6529560792550656e-05,
+      "loss": 3.9343,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.65212076026476e-05,
+      "loss": 3.9302,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.651282165513708e-05,
+      "loss": 3.9233,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.650443570762657e-05,
+      "loss": 3.9192,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.649604976011605e-05,
+      "loss": 3.9293,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.648766381260553e-05,
+      "loss": 3.9148,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6479277865095e-05,
+      "loss": 3.9291,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647089191758448e-05,
+      "loss": 3.9149,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.646250597007396e-05,
+      "loss": 3.9253,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.645413640136718e-05,
+      "loss": 3.9191,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.644575045385665e-05,
+      "loss": 3.9196,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.643736450634613e-05,
+      "loss": 3.9048,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6428994937639346e-05,
+      "loss": 3.9153,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642060899012882e-05,
+      "loss": 3.908,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6412223042618306e-05,
+      "loss": 3.9208,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6403837095107786e-05,
+      "loss": 3.9345,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6395451147597266e-05,
+      "loss": 3.926,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6387065200086746e-05,
+      "loss": 3.8975,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6378679252576226e-05,
+      "loss": 3.9278,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6370293305065705e-05,
+      "loss": 3.916,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6361923736358915e-05,
+      "loss": 3.9178,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6353537788848395e-05,
+      "loss": 3.9062,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6345151841337874e-05,
+      "loss": 3.9192,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6336765893827354e-05,
+      "loss": 3.9109,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6328396325120564e-05,
+      "loss": 3.9156,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6320010377610043e-05,
+      "loss": 3.9144,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6311624430099523e-05,
+      "loss": 3.9175,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6303238482589003e-05,
+      "loss": 3.9137,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.9084,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.62864829663717e-05,
+      "loss": 3.9175,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.627809701886118e-05,
+      "loss": 3.9149,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626971107135066e-05,
+      "loss": 3.907,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626134150264387e-05,
+      "loss": 3.9194,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.625295555513335e-05,
+      "loss": 3.905,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9699602127075195,
+      "eval_runtime": 302.1878,
+      "eval_samples_per_second": 1262.761,
+      "eval_steps_per_second": 39.462,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.8928,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623618366011231e-05,
+      "loss": 3.8991,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622779771260179e-05,
+      "loss": 3.914,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621941176509127e-05,
+      "loss": 3.9062,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621102581758075e-05,
+      "loss": 3.9192,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620263987007023e-05,
+      "loss": 3.8938,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619425392255971e-05,
+      "loss": 3.9125,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618586797504919e-05,
+      "loss": 3.8829,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617748202753867e-05,
+      "loss": 3.9144,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616909608002815e-05,
+      "loss": 3.896,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616071013251763e-05,
+      "loss": 3.9059,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615234056381084e-05,
+      "loss": 3.9047,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614397099510405e-05,
+      "loss": 3.8938,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613558504759353e-05,
+      "loss": 3.895,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127199100083006e-05,
+      "loss": 3.8889,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118813152572486e-05,
+      "loss": 3.8859,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110427205061966e-05,
+      "loss": 3.8987,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6102041257551446e-05,
+      "loss": 3.8959,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093655310040926e-05,
+      "loss": 3.8952,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085269362530406e-05,
+      "loss": 3.909,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607689979382362e-05,
+      "loss": 3.9095,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60685138463131e-05,
+      "loss": 3.9002,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606012789880258e-05,
+      "loss": 3.8948,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605175833009579e-05,
+      "loss": 3.9029,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604337238258527e-05,
+      "loss": 3.8879,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.9013,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602660048756423e-05,
+      "loss": 3.8963,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601821454005371e-05,
+      "loss": 3.8889,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600982859254319e-05,
+      "loss": 3.883,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600144264503267e-05,
+      "loss": 3.8886,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599307307632588e-05,
+      "loss": 3.8879,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5984703507619095e-05,
+      "loss": 3.8957,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.8975,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.8908,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959545665087535e-05,
+      "loss": 3.8931,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5951159717577015e-05,
+      "loss": 3.8903,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942790148870224e-05,
+      "loss": 3.8761,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.8841,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5926018253849184e-05,
+      "loss": 3.8758,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917632306338664e-05,
+      "loss": 3.8799,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909246358828144e-05,
+      "loss": 3.8819,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5900860411317624e-05,
+      "loss": 3.888,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5892474463807104e-05,
+      "loss": 3.8862,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5884088516296584e-05,
+      "loss": 3.8721,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58757189475898e-05,
+      "loss": 3.8887,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586733300007928e-05,
+      "loss": 3.8868,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585894705256876e-05,
+      "loss": 3.8857,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585057748386197e-05,
+      "loss": 3.8782,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584219153635145e-05,
+      "loss": 3.8696,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583382196764466e-05,
+      "loss": 3.8868,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582543602013414e-05,
+      "loss": 3.8738,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581705007262362e-05,
+      "loss": 3.8798,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58086641251131e-05,
+      "loss": 3.8713,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580027817760258e-05,
+      "loss": 3.8739,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579190860889579e-05,
+      "loss": 3.8664,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783522661385267e-05,
+      "loss": 3.8759,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577513671387475e-05,
+      "loss": 3.8757,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576675076636423e-05,
+      "loss": 3.8763,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575836481885371e-05,
+      "loss": 3.8776,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574997887134319e-05,
+      "loss": 3.8664,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741592923832666e-05,
+      "loss": 3.8664,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733206976322146e-05,
+      "loss": 3.8857,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572485378641909e-05,
+      "loss": 3.8567,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571646783890857e-05,
+      "loss": 3.8553,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570808189139805e-05,
+      "loss": 3.8673,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569969594388753e-05,
+      "loss": 3.877,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569130999637701e-05,
+      "loss": 3.8654,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568292404886649e-05,
+      "loss": 3.8615,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567453810135597e-05,
+      "loss": 3.8559,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566615215384545e-05,
+      "loss": 3.862,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565778258513867e-05,
+      "loss": 3.8759,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.8657,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5641027068921356e-05,
+      "loss": 3.8662,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5632641121410836e-05,
+      "loss": 3.8762,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5624255173900316e-05,
+      "loss": 3.8752,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561586922638979e-05,
+      "loss": 3.8578,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560748327887927e-05,
+      "loss": 3.8625,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559909733136875e-05,
+      "loss": 3.8711,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559071138385823e-05,
+      "loss": 3.8511,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5582358193955174e-05,
+      "loss": 3.8679,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5573972246444654e-05,
+      "loss": 3.8557,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556558629893414e-05,
+      "loss": 3.8773,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557200351423614e-05,
+      "loss": 3.8612,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548814403913094e-05,
+      "loss": 3.8528,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540428456402574e-05,
+      "loss": 3.8508,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553205888769579e-05,
+      "loss": 3.8631,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552367294018526e-05,
+      "loss": 3.874,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551528699267474e-05,
+      "loss": 3.857,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550690104516422e-05,
+      "loss": 3.8545,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54985150976537e-05,
+      "loss": 3.8541,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549012915014318e-05,
+      "loss": 3.8651,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548174320263266e-05,
+      "loss": 3.8492,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547335725512214e-05,
+      "loss": 3.8543,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546500406521909e-05,
+      "loss": 3.8625,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54566344965123e-05,
+      "loss": 3.8668,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544824854900178e-05,
+      "loss": 3.8615,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543986260149126e-05,
+      "loss": 3.8442,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5431476653980736e-05,
+      "loss": 3.8554,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5423090706470216e-05,
+      "loss": 3.8486,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5414704758959696e-05,
+      "loss": 3.8578,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406318811449176e-05,
+      "loss": 3.8592,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5397932863938656e-05,
+      "loss": 3.847,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389546916428136e-05,
+      "loss": 3.8479,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5381160968917616e-05,
+      "loss": 3.856,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372775021407096e-05,
+      "loss": 3.8341,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536440545270031e-05,
+      "loss": 3.8505,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535601950518979e-05,
+      "loss": 3.8396,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534763355767927e-05,
+      "loss": 3.8578,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533924761016875e-05,
+      "loss": 3.8473,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533087804146196e-05,
+      "loss": 3.8477,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532249209395144e-05,
+      "loss": 3.8425,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531410614644092e-05,
+      "loss": 3.8507,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53057201989304e-05,
+      "loss": 3.8543,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529733425141988e-05,
+      "loss": 3.8437,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528894830390936e-05,
+      "loss": 3.8525,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528057873520257e-05,
+      "loss": 3.857,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527219278769205e-05,
+      "loss": 3.8492,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526380684018153e-05,
+      "loss": 3.8529,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255420892671017e-05,
+      "loss": 3.8377,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5247051323964226e-05,
+      "loss": 3.8549,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238665376453706e-05,
+      "loss": 3.844,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5230279428943186e-05,
+      "loss": 3.8492,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5221893481432665e-05,
+      "loss": 3.8393,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5213523912725875e-05,
+      "loss": 3.8481,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5205137965215355e-05,
+      "loss": 3.8424,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5196752017704834e-05,
+      "loss": 3.8409,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5188366070194314e-05,
+      "loss": 3.8319,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5179996501487524e-05,
+      "loss": 3.8358,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5171610553977003e-05,
+      "loss": 3.8387,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5163224606466483e-05,
+      "loss": 3.8438,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515483865895597e-05,
+      "loss": 3.8574,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514645271144545e-05,
+      "loss": 3.8515,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513806676393492e-05,
+      "loss": 3.8246,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51296808164244e-05,
+      "loss": 3.854,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512129486891388e-05,
+      "loss": 3.8406,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51129253002071e-05,
+      "loss": 3.8421,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510453935269657e-05,
+      "loss": 3.8337,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509615340518605e-05,
+      "loss": 3.8452,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508778383647927e-05,
+      "loss": 3.8362,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507939788896874e-05,
+      "loss": 3.8481,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507101194145822e-05,
+      "loss": 3.8395,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506262599394771e-05,
+      "loss": 3.8411,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054256425240924e-05,
+      "loss": 3.845,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50458704777304e-05,
+      "loss": 3.8337,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503748453021988e-05,
+      "loss": 3.8423,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502909858270936e-05,
+      "loss": 3.8473,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502071263519884e-05,
+      "loss": 3.8336,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501232668768832e-05,
+      "loss": 3.8419,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5003957118981526e-05,
+      "loss": 3.8377,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9284117221832275,
+      "eval_runtime": 330.261,
+      "eval_samples_per_second": 1155.422,
+      "eval_steps_per_second": 36.108,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995571171471006e-05,
+      "loss": 3.8281,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4987185223960486e-05,
+      "loss": 3.8272,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978799276449966e-05,
+      "loss": 3.8388,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970413328939446e-05,
+      "loss": 3.8385,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496204376023266e-05,
+      "loss": 3.8515,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.495365781272214e-05,
+      "loss": 3.8208,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494527186521162e-05,
+      "loss": 3.8376,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49368859177011e-05,
+      "loss": 3.8186,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492849997019058e-05,
+      "loss": 3.8393,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492013040148379e-05,
+      "loss": 3.8259,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491174445397327e-05,
+      "loss": 3.8321,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490335850646275e-05,
+      "loss": 3.8387,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489498893775596e-05,
+      "loss": 3.8213,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488660299024544e-05,
+      "loss": 3.8285,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487821704273492e-05,
+      "loss": 3.82,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48698310952244e-05,
+      "loss": 3.8174,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486144514771388e-05,
+      "loss": 3.8269,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485305920020336e-05,
+      "loss": 3.8286,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.484467325269284e-05,
+      "loss": 3.8249,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4836287305182326e-05,
+      "loss": 3.8407,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4827901357671806e-05,
+      "loss": 3.8414,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4819531788965015e-05,
+      "loss": 3.8322,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4811145841454495e-05,
+      "loss": 3.8282,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4802759893943975e-05,
+      "loss": 3.8364,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4794373946433455e-05,
+      "loss": 3.8173,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4785987998922935e-05,
+      "loss": 3.8334,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4777618430216144e-05,
+      "loss": 3.8279,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4769232482705624e-05,
+      "loss": 3.8199,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4760846535195104e-05,
+      "loss": 3.817,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475246058768458e-05,
+      "loss": 3.8182,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4744074640174064e-05,
+      "loss": 3.8239,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473570507146728e-05,
+      "loss": 3.8282,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472731912395676e-05,
+      "loss": 3.8315,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471893317644623e-05,
+      "loss": 3.8252,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471054722893571e-05,
+      "loss": 3.8253,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470217766022893e-05,
+      "loss": 3.8252,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469379171271841e-05,
+      "loss": 3.8087,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468540576520788e-05,
+      "loss": 3.8215,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467701981769736e-05,
+      "loss": 3.8098,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466865024899058e-05,
+      "loss": 3.8137,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466026430148005e-05,
+      "loss": 3.8136,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465187835396953e-05,
+      "loss": 3.8276,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464349240645902e-05,
+      "loss": 3.8187,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.463512283775223e-05,
+      "loss": 3.807,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626736890241707e-05,
+      "loss": 3.8252,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618350942731186e-05,
+      "loss": 3.82,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609964995220666e-05,
+      "loss": 3.8192,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460159542651388e-05,
+      "loss": 3.8161,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4593209479003356e-05,
+      "loss": 3.8037,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584823531492835e-05,
+      "loss": 3.8245,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576437583982315e-05,
+      "loss": 3.8082,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4568051636471795e-05,
+      "loss": 3.819,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559682067765004e-05,
+      "loss": 3.8078,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551296120254484e-05,
+      "loss": 3.808,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454291017274397e-05,
+      "loss": 3.8047,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453452422523345e-05,
+      "loss": 3.8129,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452615465652666e-05,
+      "loss": 3.8111,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451776870901614e-05,
+      "loss": 3.8125,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450938276150562e-05,
+      "loss": 3.8176,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45009968139951e-05,
+      "loss": 3.8042,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449261086648458e-05,
+      "loss": 3.7997,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448424129777779e-05,
+      "loss": 3.8267,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447585535026727e-05,
+      "loss": 3.792,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446746940275675e-05,
+      "loss": 3.7934,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445908345524623e-05,
+      "loss": 3.8051,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445071388653944e-05,
+      "loss": 3.8193,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4442327939028925e-05,
+      "loss": 3.798,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4433941991518405e-05,
+      "loss": 3.803,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425556044007885e-05,
+      "loss": 3.7941,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4417170096497365e-05,
+      "loss": 3.8001,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4408784148986845e-05,
+      "loss": 3.8114,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4400414580280054e-05,
+      "loss": 3.8082,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4392028632769534e-05,
+      "loss": 3.8064,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4383642685259014e-05,
+      "loss": 3.8118,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4375256737748494e-05,
+      "loss": 3.8138,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43668871690417e-05,
+      "loss": 3.8,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435850122153118e-05,
+      "loss": 3.8011,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435011527402066e-05,
+      "loss": 3.8105,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434172932651014e-05,
+      "loss": 3.7891,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433335975780336e-05,
+      "loss": 3.8061,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432497381029284e-05,
+      "loss": 3.7963,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431658786278232e-05,
+      "loss": 3.8172,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43082019152718e-05,
+      "loss": 3.801,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429983234656501e-05,
+      "loss": 3.795,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429144639905449e-05,
+      "loss": 3.7848,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428306045154397e-05,
+      "loss": 3.8071,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427467450403345e-05,
+      "loss": 3.8109,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266304935326656e-05,
+      "loss": 3.8021,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4257918987816136e-05,
+      "loss": 3.7928,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4249533040305616e-05,
+      "loss": 3.7966,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4241147092795096e-05,
+      "loss": 3.8038,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423277752408831e-05,
+      "loss": 3.7912,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422439157657779e-05,
+      "loss": 3.7926,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421600562906727e-05,
+      "loss": 3.8046,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.420763606036048e-05,
+      "loss": 3.8066,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419925011284996e-05,
+      "loss": 3.8022,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419086416533944e-05,
+      "loss": 3.7882,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.418247821782892e-05,
+      "loss": 3.7973,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41740922703184e-05,
+      "loss": 3.789,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.416570632280788e-05,
+      "loss": 3.7987,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.415732037529736e-05,
+      "loss": 3.7984,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148934427786834e-05,
+      "loss": 3.7934,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4140548480276314e-05,
+      "loss": 3.7899,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.413217891156953e-05,
+      "loss": 3.7923,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412379296405901e-05,
+      "loss": 3.7857,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411540701654849e-05,
+      "loss": 3.7877,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410702106903797e-05,
+      "loss": 3.7818,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4098651500331186e-05,
+      "loss": 3.8003,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409026555282066e-05,
+      "loss": 3.7915,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408187960531014e-05,
+      "loss": 3.788,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407349365779962e-05,
+      "loss": 3.7835,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4065124089092835e-05,
+      "loss": 3.796,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405673814158231e-05,
+      "loss": 3.7966,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404835219407179e-05,
+      "loss": 3.7861,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403996624656127e-05,
+      "loss": 3.7961,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4031596677854484e-05,
+      "loss": 3.8009,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4023210730343963e-05,
+      "loss": 3.7927,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4014824782833443e-05,
+      "loss": 3.7975,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4006438835322923e-05,
+      "loss": 3.7829,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399806926661613e-05,
+      "loss": 3.7944,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398968331910561e-05,
+      "loss": 3.7902,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398129737159509e-05,
+      "loss": 3.7941,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397291142408457e-05,
+      "loss": 3.779,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396454185537778e-05,
+      "loss": 3.793,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395615590786726e-05,
+      "loss": 3.7872,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394776996035674e-05,
+      "loss": 3.784,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393938401284622e-05,
+      "loss": 3.7781,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393101444413944e-05,
+      "loss": 3.7809,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392262849662892e-05,
+      "loss": 3.7808,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39142425491184e-05,
+      "loss": 3.7922,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390585660160788e-05,
+      "loss": 3.7981,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3897487032901086e-05,
+      "loss": 3.8003,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3889101085390566e-05,
+      "loss": 3.7699,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3880715137880046e-05,
+      "loss": 3.7962,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3872329190369526e-05,
+      "loss": 3.7888,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3863959621662735e-05,
+      "loss": 3.7822,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3855573674152215e-05,
+      "loss": 3.7796,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3847187726641695e-05,
+      "loss": 3.7945,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3838801779131175e-05,
+      "loss": 3.7831,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383043221042439e-05,
+      "loss": 3.7914,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382204626291387e-05,
+      "loss": 3.7818,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381366031540335e-05,
+      "loss": 3.7854,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.380527436789283e-05,
+      "loss": 3.7908,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379690479918604e-05,
+      "loss": 3.7814,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378851885167552e-05,
+      "loss": 3.7862,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3780132904165e-05,
+      "loss": 3.7946,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377174695665448e-05,
+      "loss": 3.7794,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.376336100914396e-05,
+      "loss": 3.7829,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375499144043717e-05,
+      "loss": 3.7866,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9028751850128174,
+      "eval_runtime": 335.3025,
+      "eval_samples_per_second": 1138.05,
+      "eval_steps_per_second": 35.565,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374660549292665e-05,
+      "loss": 3.7705,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373821954541613e-05,
+      "loss": 3.772,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372983359790561e-05,
+      "loss": 3.7865,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721447650395095e-05,
+      "loss": 3.787,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3713061702884575e-05,
+      "loss": 3.7931,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704675755374055e-05,
+      "loss": 3.7751,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696289807863535e-05,
+      "loss": 3.782,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687903860353015e-05,
+      "loss": 3.7665,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3679534291646224e-05,
+      "loss": 3.7831,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3671148344135704e-05,
+      "loss": 3.7775,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3662762396625184e-05,
+      "loss": 3.7773,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3654376449114664e-05,
+      "loss": 3.784,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36460232592116e-05,
+      "loss": 3.7717,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363763731170108e-05,
+      "loss": 3.7736,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362925136419056e-05,
+      "loss": 3.7704,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362086541668005e-05,
+      "loss": 3.7645,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361247946916953e-05,
+      "loss": 3.7707,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360409352165901e-05,
+      "loss": 3.7768,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359570757414849e-05,
+      "loss": 3.7733,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358732162663797e-05,
+      "loss": 3.7896,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357895205793118e-05,
+      "loss": 3.787,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357056611042066e-05,
+      "loss": 3.7819,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356218016291014e-05,
+      "loss": 3.777,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355379421539962e-05,
+      "loss": 3.7835,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354542464669283e-05,
+      "loss": 3.7672,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353703869918231e-05,
+      "loss": 3.7824,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352865275167179e-05,
+      "loss": 3.7757,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352026680416127e-05,
+      "loss": 3.7668,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351189723545448e-05,
+      "loss": 3.7659,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350351128794396e-05,
+      "loss": 3.7706,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349512534043344e-05,
+      "loss": 3.7723,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486739392922916e-05,
+      "loss": 3.7755,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347836982421613e-05,
+      "loss": 3.7832,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346998387670561e-05,
+      "loss": 3.7743,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346159792919509e-05,
+      "loss": 3.7748,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453211981684565e-05,
+      "loss": 3.7754,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344484241297778e-05,
+      "loss": 3.762,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343645646546726e-05,
+      "loss": 3.7643,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342807051795674e-05,
+      "loss": 3.7605,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341968457044622e-05,
+      "loss": 3.7645,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411315001739436e-05,
+      "loss": 3.7631,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402929054228916e-05,
+      "loss": 3.7782,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339454310671839e-05,
+      "loss": 3.7696,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338615715920787e-05,
+      "loss": 3.7567,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377787590501085e-05,
+      "loss": 3.7776,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369401642990565e-05,
+      "loss": 3.7665,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336101569548004e-05,
+      "loss": 3.7741,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335262974796952e-05,
+      "loss": 3.7632,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3344243800459e-05,
+      "loss": 3.754,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333587423175221e-05,
+      "loss": 3.7717,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332748828424169e-05,
+      "loss": 3.7624,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3319102336731174e-05,
+      "loss": 3.7709,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3310716389220654e-05,
+      "loss": 3.754,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330234682051386e-05,
+      "loss": 3.7616,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329396087300334e-05,
+      "loss": 3.7516,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328557492549282e-05,
+      "loss": 3.764,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32771889779823e-05,
+      "loss": 3.7653,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326881940927551e-05,
+      "loss": 3.7597,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326043346176499e-05,
+      "loss": 3.7706,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325204751425447e-05,
+      "loss": 3.7573,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324366156674395e-05,
+      "loss": 3.7496,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323529199803716e-05,
+      "loss": 3.7807,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322690605052664e-05,
+      "loss": 3.7441,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321852010301613e-05,
+      "loss": 3.7397,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321013415550561e-05,
+      "loss": 3.7573,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320176458679882e-05,
+      "loss": 3.7717,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31933786392883e-05,
+      "loss": 3.7456,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318499269177778e-05,
+      "loss": 3.7568,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176623123070986e-05,
+      "loss": 3.7452,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3168237175560466e-05,
+      "loss": 3.7513,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159851228049946e-05,
+      "loss": 3.7654,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151465280539426e-05,
+      "loss": 3.761,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143095711832635e-05,
+      "loss": 3.7599,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134709764322115e-05,
+      "loss": 3.76,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126323816811595e-05,
+      "loss": 3.7686,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311793786930108e-05,
+      "loss": 3.7494,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310956830059429e-05,
+      "loss": 3.7584,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310118235308377e-05,
+      "loss": 3.7603,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309279640557325e-05,
+      "loss": 3.7426,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308441045806273e-05,
+      "loss": 3.7566,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307604088935594e-05,
+      "loss": 3.7532,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306765494184542e-05,
+      "loss": 3.7661,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30592689943349e-05,
+      "loss": 3.756,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305088304682438e-05,
+      "loss": 3.7482,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304251347811759e-05,
+      "loss": 3.7393,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303412753060707e-05,
+      "loss": 3.7565,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302574158309655e-05,
+      "loss": 3.7626,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017355635586035e-05,
+      "loss": 3.7532,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008986066879244e-05,
+      "loss": 3.7513,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000600119368724e-05,
+      "loss": 3.7516,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2992214171858204e-05,
+      "loss": 3.7558,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983828224347684e-05,
+      "loss": 3.7454,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297545865564089e-05,
+      "loss": 3.7466,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296707270813037e-05,
+      "loss": 3.7592,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295870313942358e-05,
+      "loss": 3.7587,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295031719191306e-05,
+      "loss": 3.7555,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294193124440254e-05,
+      "loss": 3.7411,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293354529689202e-05,
+      "loss": 3.7541,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29251593493815e-05,
+      "loss": 3.7393,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291677340187099e-05,
+      "loss": 3.7547,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290838745436047e-05,
+      "loss": 3.7513,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290000150684995e-05,
+      "loss": 3.7491,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289163193814316e-05,
+      "loss": 3.7444,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288324599063264e-05,
+      "loss": 3.7408,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287486004312212e-05,
+      "loss": 3.7472,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28664740956116e-05,
+      "loss": 3.7386,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285810452690481e-05,
+      "loss": 3.7402,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284971857939429e-05,
+      "loss": 3.7515,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284133263188377e-05,
+      "loss": 3.7501,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2832946684373247e-05,
+      "loss": 3.7439,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824577115666456e-05,
+      "loss": 3.7384,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281619116815594e-05,
+      "loss": 3.7482,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280780522064542e-05,
+      "loss": 3.7519,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27994192731349e-05,
+      "loss": 3.7466,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279104970442811e-05,
+      "loss": 3.7526,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278266375691759e-05,
+      "loss": 3.7541,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277427780940707e-05,
+      "loss": 3.7512,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276589186189655e-05,
+      "loss": 3.7497,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275752229318976e-05,
+      "loss": 3.7425,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274913634567924e-05,
+      "loss": 3.7452,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274075039816872e-05,
+      "loss": 3.7476,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27323644506582e-05,
+      "loss": 3.75,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272399488195141e-05,
+      "loss": 3.7368,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715608934440896e-05,
+      "loss": 3.7465,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707222986930376e-05,
+      "loss": 3.7451,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698837039419856e-05,
+      "loss": 3.741,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690467470713065e-05,
+      "loss": 3.732,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2682081523202545e-05,
+      "loss": 3.7405,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673695575692025e-05,
+      "loss": 3.7357,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2665309628181505e-05,
+      "loss": 3.7479,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656940059474714e-05,
+      "loss": 3.7512,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648554111964194e-05,
+      "loss": 3.7574,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2640168164453674e-05,
+      "loss": 3.7271,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2631782216943154e-05,
+      "loss": 3.7535,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262341264823636e-05,
+      "loss": 3.7463,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261502670072584e-05,
+      "loss": 3.7397,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260664075321533e-05,
+      "loss": 3.735,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259825480570481e-05,
+      "loss": 3.752,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258988523699802e-05,
+      "loss": 3.736,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25814992894875e-05,
+      "loss": 3.7511,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257311334197698e-05,
+      "loss": 3.7353,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256472739446646e-05,
+      "loss": 3.7457,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255635782575967e-05,
+      "loss": 3.7448,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254797187824915e-05,
+      "loss": 3.7379,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253958593073863e-05,
+      "loss": 3.7491,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253119998322811e-05,
+      "loss": 3.7437,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252283041452132e-05,
+      "loss": 3.7466,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25144444670108e-05,
+      "loss": 3.7307,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250605851950028e-05,
+      "loss": 3.7487,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.887489080429077,
+      "eval_runtime": 303.274,
+      "eval_samples_per_second": 1258.238,
+      "eval_steps_per_second": 39.321,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249767257198976e-05,
+      "loss": 3.7367,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248930300328297e-05,
+      "loss": 3.7284,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248091705577245e-05,
+      "loss": 3.7384,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247253110826193e-05,
+      "loss": 3.7478,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246414516075141e-05,
+      "loss": 3.7507,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245577559204462e-05,
+      "loss": 3.7313,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24473896445341e-05,
+      "loss": 3.7377,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243900369702358e-05,
+      "loss": 3.7249,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2430617749513054e-05,
+      "loss": 3.7424,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242224818080627e-05,
+      "loss": 3.7336,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241386223329575e-05,
+      "loss": 3.7383,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240547628578524e-05,
+      "loss": 3.7401,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2397106717078446e-05,
+      "loss": 3.7305,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2388720769567926e-05,
+      "loss": 3.7347,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2380334822057406e-05,
+      "loss": 3.7247,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2371948874546886e-05,
+      "loss": 3.7214,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236356292703636e-05,
+      "loss": 3.7296,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235517697952584e-05,
+      "loss": 3.7407,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234679103201532e-05,
+      "loss": 3.7298,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23384050845048e-05,
+      "loss": 3.7518,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233003551579801e-05,
+      "loss": 3.743,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232164956828749e-05,
+      "loss": 3.7421,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2313263620776975e-05,
+      "loss": 3.7353,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2304877673266455e-05,
+      "loss": 3.7379,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2296508104559664e-05,
+      "loss": 3.7248,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2288122157049144e-05,
+      "loss": 3.7447,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2279736209538624e-05,
+      "loss": 3.7315,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2271350262028104e-05,
+      "loss": 3.7255,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226298069332131e-05,
+      "loss": 3.7281,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225459474581079e-05,
+      "loss": 3.7296,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224620879830027e-05,
+      "loss": 3.7306,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223782285078975e-05,
+      "loss": 3.7375,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222945328208296e-05,
+      "loss": 3.7372,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222106733457244e-05,
+      "loss": 3.739,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221268138706193e-05,
+      "loss": 3.7281,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220431181835514e-05,
+      "loss": 3.7338,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219592587084462e-05,
+      "loss": 3.7286,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21875399233341e-05,
+      "loss": 3.7199,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217915397582358e-05,
+      "loss": 3.7188,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2170784407116787e-05,
+      "loss": 3.723,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2162398459606267e-05,
+      "loss": 3.7239,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2154012512095746e-05,
+      "loss": 3.7382,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2145626564585226e-05,
+      "loss": 3.7274,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2137256995878436e-05,
+      "loss": 3.7197,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128871048367915e-05,
+      "loss": 3.7397,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2120485100857395e-05,
+      "loss": 3.724,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211209915334688e-05,
+      "loss": 3.7345,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210372958464009e-05,
+      "loss": 3.7239,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209534363712957e-05,
+      "loss": 3.7079,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208695768961905e-05,
+      "loss": 3.7386,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207857174210853e-05,
+      "loss": 3.72,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207020217340174e-05,
+      "loss": 3.7318,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206181622589122e-05,
+      "loss": 3.7197,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20534302783807e-05,
+      "loss": 3.7214,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204504433087018e-05,
+      "loss": 3.7085,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.203667476216339e-05,
+      "loss": 3.721,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.202828881465287e-05,
+      "loss": 3.7309,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201990286714235e-05,
+      "loss": 3.7167,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011516919631836e-05,
+      "loss": 3.7317,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2003147350925045e-05,
+      "loss": 3.7223,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1994761403414525e-05,
+      "loss": 3.7112,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1986375455904005e-05,
+      "loss": 3.7363,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1977989508393485e-05,
+      "loss": 3.7075,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1969619939686694e-05,
+      "loss": 3.703,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1961233992176174e-05,
+      "loss": 3.7214,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1952848044665654e-05,
+      "loss": 3.7291,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1944462097155134e-05,
+      "loss": 3.7069,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193609252844834e-05,
+      "loss": 3.7217,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192770658093782e-05,
+      "loss": 3.7041,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19193206334273e-05,
+      "loss": 3.7177,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191093468591679e-05,
+      "loss": 3.7224,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190256511721e-05,
+      "loss": 3.7193,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189417916969948e-05,
+      "loss": 3.723,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188579322218896e-05,
+      "loss": 3.7185,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187740727467844e-05,
+      "loss": 3.7308,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186903770597165e-05,
+      "loss": 3.7104,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186065175846113e-05,
+      "loss": 3.7214,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185226581095061e-05,
+      "loss": 3.7173,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184387986344009e-05,
+      "loss": 3.7056,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835510294733297e-05,
+      "loss": 3.7188,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1827124347222776e-05,
+      "loss": 3.7152,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818738399712256e-05,
+      "loss": 3.7278,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181035245220174e-05,
+      "loss": 3.723,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180198288349495e-05,
+      "loss": 3.7051,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179359693598443e-05,
+      "loss": 3.7026,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178521098847391e-05,
+      "loss": 3.7187,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177682504096339e-05,
+      "loss": 3.7272,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.17684554722566e-05,
+      "loss": 3.7165,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176006952474608e-05,
+      "loss": 3.7081,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175168357723556e-05,
+      "loss": 3.7163,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174329762972504e-05,
+      "loss": 3.7145,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173492806101825e-05,
+      "loss": 3.7107,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172654211350773e-05,
+      "loss": 3.7061,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171815616599721e-05,
+      "loss": 3.7202,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1709786597290426e-05,
+      "loss": 3.7187,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1701400649779906e-05,
+      "loss": 3.7199,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1693014702269386e-05,
+      "loss": 3.706,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1684628754758866e-05,
+      "loss": 3.7108,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1676242807248346e-05,
+      "loss": 3.7068,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1667856859737826e-05,
+      "loss": 3.7172,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1659470912227306e-05,
+      "loss": 3.7132,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1651084964716786e-05,
+      "loss": 3.7095,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642715396009995e-05,
+      "loss": 3.7117,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1634329448499475e-05,
+      "loss": 3.7047,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1625943500988955e-05,
+      "loss": 3.7089,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1617557553478435e-05,
+      "loss": 3.7014,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1609187984771644e-05,
+      "loss": 3.7073,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160080203726113e-05,
+      "loss": 3.708,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159241608975061e-05,
+      "loss": 3.7177,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1584030142240084e-05,
+      "loss": 3.7018,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15756605735333e-05,
+      "loss": 3.7046,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156727462602278e-05,
+      "loss": 3.7068,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155888867851226e-05,
+      "loss": 3.7138,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155050273100173e-05,
+      "loss": 3.7087,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154213316229495e-05,
+      "loss": 3.719,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153374721478443e-05,
+      "loss": 3.7136,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152536126727391e-05,
+      "loss": 3.7141,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151697531976338e-05,
+      "loss": 3.7116,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15086057510566e-05,
+      "loss": 3.7095,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1500219803546084e-05,
+      "loss": 3.7083,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149183385603556e-05,
+      "loss": 3.7117,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148344790852504e-05,
+      "loss": 3.7087,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147507833981825e-05,
+      "loss": 3.7029,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146669239230773e-05,
+      "loss": 3.7128,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1458306444797206e-05,
+      "loss": 3.7083,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1449920497286686e-05,
+      "loss": 3.705,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14415509285799e-05,
+      "loss": 3.6937,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143316498106938e-05,
+      "loss": 3.7034,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1424779033558855e-05,
+      "loss": 3.7007,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1416393086048335e-05,
+      "loss": 3.7096,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140802351734155e-05,
+      "loss": 3.7159,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139963756983103e-05,
+      "loss": 3.7197,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139125162232051e-05,
+      "loss": 3.6922,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138286567480999e-05,
+      "loss": 3.7144,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.137449610610321e-05,
+      "loss": 3.7079,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136611015859268e-05,
+      "loss": 3.7086,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.135772421108216e-05,
+      "loss": 3.6985,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1349354642375376e-05,
+      "loss": 3.7143,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340968694864856e-05,
+      "loss": 3.6982,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133258274735433e-05,
+      "loss": 3.7166,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132419679984381e-05,
+      "loss": 3.6974,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1315827231137025e-05,
+      "loss": 3.7093,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1307441283626505e-05,
+      "loss": 3.7093,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1299055336115985e-05,
+      "loss": 3.7061,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1290669388605465e-05,
+      "loss": 3.7105,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128229981989868e-05,
+      "loss": 3.7082,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1273913872388154e-05,
+      "loss": 3.7088,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1265527924877634e-05,
+      "loss": 3.6967,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1257141977367114e-05,
+      "loss": 3.7143,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.877584934234619,
+      "eval_runtime": 302.778,
+      "eval_samples_per_second": 1260.3,
+      "eval_steps_per_second": 39.385,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248756029856594e-05,
+      "loss": 3.6994,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1240370082346074e-05,
+      "loss": 3.6918,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1231984134835553e-05,
+      "loss": 3.707,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1223598187325033e-05,
+      "loss": 3.7097,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1215212239814513e-05,
+      "loss": 3.7138,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120682629230399e-05,
+      "loss": 3.6967,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119844034479347e-05,
+      "loss": 3.7052,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119005439728296e-05,
+      "loss": 3.6903,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118168482857617e-05,
+      "loss": 3.7045,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117329888106565e-05,
+      "loss": 3.7003,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116491293355513e-05,
+      "loss": 3.701,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115652698604461e-05,
+      "loss": 3.7042,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114817379614155e-05,
+      "loss": 3.6993,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113978784863103e-05,
+      "loss": 3.6951,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113140190112051e-05,
+      "loss": 3.6917,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112301595360999e-05,
+      "loss": 3.6859,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111463000609947e-05,
+      "loss": 3.695,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110624405858895e-05,
+      "loss": 3.7032,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109785811107843e-05,
+      "loss": 3.6981,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108947216356791e-05,
+      "loss": 3.7095,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108110259486112e-05,
+      "loss": 3.7114,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10727166473506e-05,
+      "loss": 3.7097,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106433069984008e-05,
+      "loss": 3.6992,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105594475232956e-05,
+      "loss": 3.6994,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104757518362277e-05,
+      "loss": 3.6971,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103918923611225e-05,
+      "loss": 3.7069,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103080328860173e-05,
+      "loss": 3.6948,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102241734109121e-05,
+      "loss": 3.6959,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101404777238442e-05,
+      "loss": 3.6933,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10056618248739e-05,
+      "loss": 3.6934,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099727587736338e-05,
+      "loss": 3.6917,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098888992985286e-05,
+      "loss": 3.7044,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0980520361146076e-05,
+      "loss": 3.7028,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0972134413635556e-05,
+      "loss": 3.7049,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963748466125036e-05,
+      "loss": 3.6909,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0955378897418245e-05,
+      "loss": 3.703,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946992949907725e-05,
+      "loss": 3.6984,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938607002397205e-05,
+      "loss": 3.6832,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930221054886685e-05,
+      "loss": 3.686,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921851486179894e-05,
+      "loss": 3.6881,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913465538669374e-05,
+      "loss": 3.694,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905079591158854e-05,
+      "loss": 3.6983,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896693643648334e-05,
+      "loss": 3.6979,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088832407494155e-05,
+      "loss": 3.6836,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087993812743103e-05,
+      "loss": 3.7051,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087155217992051e-05,
+      "loss": 3.692,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086316623240999e-05,
+      "loss": 3.7018,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08547966637032e-05,
+      "loss": 3.693,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084641071619268e-05,
+      "loss": 3.6696,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083802476868216e-05,
+      "loss": 3.7081,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082963882117163e-05,
+      "loss": 3.6851,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082126925246485e-05,
+      "loss": 3.6971,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081288330495433e-05,
+      "loss": 3.6893,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080449735744381e-05,
+      "loss": 3.6871,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079611140993329e-05,
+      "loss": 3.6802,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0787741841226504e-05,
+      "loss": 3.6846,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0779355893715984e-05,
+      "loss": 3.7021,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077098632500919e-05,
+      "loss": 3.6794,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076260037749867e-05,
+      "loss": 3.697,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075421442998815e-05,
+      "loss": 3.6938,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074582848247763e-05,
+      "loss": 3.6761,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737442534967106e-05,
+      "loss": 3.7017,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0729056587456586e-05,
+      "loss": 3.6787,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720670639946066e-05,
+      "loss": 3.6713,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0712284692435546e-05,
+      "loss": 3.685,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070391512372876e-05,
+      "loss": 3.6931,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069552917621824e-05,
+      "loss": 3.6745,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068714322870772e-05,
+      "loss": 3.6884,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06787572811972e-05,
+      "loss": 3.6729,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067038771249041e-05,
+      "loss": 3.6818,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066200176497989e-05,
+      "loss": 3.6914,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065361581746937e-05,
+      "loss": 3.6844,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064522986995885e-05,
+      "loss": 3.6949,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063686030125206e-05,
+      "loss": 3.6849,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062847435374154e-05,
+      "loss": 3.6951,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062008840623102e-05,
+      "loss": 3.6832,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06117024587205e-05,
+      "loss": 3.6905,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0603332890013715e-05,
+      "loss": 3.6794,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594946942503195e-05,
+      "loss": 3.6765,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586560994992675e-05,
+      "loss": 3.6858,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578175047482155e-05,
+      "loss": 3.6833,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569805478775364e-05,
+      "loss": 3.6929,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0561419531264844e-05,
+      "loss": 3.6951,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0553033583754324e-05,
+      "loss": 3.6696,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0544647636243804e-05,
+      "loss": 3.6708,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053627806753701e-05,
+      "loss": 3.6847,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052789212002649e-05,
+      "loss": 3.6935,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051950617251597e-05,
+      "loss": 3.6837,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051112022500545e-05,
+      "loss": 3.6744,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050275065629867e-05,
+      "loss": 3.6887,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049436470878815e-05,
+      "loss": 3.6806,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048597876127763e-05,
+      "loss": 3.679,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047759281376711e-05,
+      "loss": 3.6733,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046922324506032e-05,
+      "loss": 3.6865,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046085367635353e-05,
+      "loss": 3.6903,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045246772884301e-05,
+      "loss": 3.6891,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044408178133249e-05,
+      "loss": 3.674,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043569583382197e-05,
+      "loss": 3.6785,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042730988631145e-05,
+      "loss": 3.6746,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041892393880093e-05,
+      "loss": 3.6861,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041053799129041e-05,
+      "loss": 3.6829,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040215204377989e-05,
+      "loss": 3.677,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03937824750731e-05,
+      "loss": 3.6785,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038539652756258e-05,
+      "loss": 3.6739,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037701058005206e-05,
+      "loss": 3.6748,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036862463254154e-05,
+      "loss": 3.6686,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036025506383475e-05,
+      "loss": 3.6772,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035186911632423e-05,
+      "loss": 3.6765,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034348316881371e-05,
+      "loss": 3.684,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033509722130319e-05,
+      "loss": 3.6735,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03267276525964e-05,
+      "loss": 3.6666,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031834170508588e-05,
+      "loss": 3.6781,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030995575757536e-05,
+      "loss": 3.6807,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030156981006484e-05,
+      "loss": 3.6806,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0293200241358056e-05,
+      "loss": 3.6851,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284814293847536e-05,
+      "loss": 3.6816,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0276428346337016e-05,
+      "loss": 3.6898,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0268042398826496e-05,
+      "loss": 3.6795,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259672830119705e-05,
+      "loss": 3.6793,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0251286882609185e-05,
+      "loss": 3.6744,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242900935098665e-05,
+      "loss": 3.687,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234514987588145e-05,
+      "loss": 3.6771,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0226145418881354e-05,
+      "loss": 3.6701,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217759471370834e-05,
+      "loss": 3.6795,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020938990266404e-05,
+      "loss": 3.6826,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020100395515353e-05,
+      "loss": 3.6709,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019261800764301e-05,
+      "loss": 3.6659,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018423206013249e-05,
+      "loss": 3.6695,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017584611262197e-05,
+      "loss": 3.6704,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016746016511145e-05,
+      "loss": 3.6769,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015907421760093e-05,
+      "loss": 3.6844,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015070464889414e-05,
+      "loss": 3.6882,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014231870138362e-05,
+      "loss": 3.6611,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01339327538731e-05,
+      "loss": 3.6863,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012554680636258e-05,
+      "loss": 3.677,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011716085885206e-05,
+      "loss": 3.6744,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010877491134154e-05,
+      "loss": 3.6682,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010038896383101e-05,
+      "loss": 3.6807,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009200301632049e-05,
+      "loss": 3.6701,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008363344761371e-05,
+      "loss": 3.6899,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007524750010319e-05,
+      "loss": 3.6672,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006686155259267e-05,
+      "loss": 3.676,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0058491983885883e-05,
+      "loss": 3.6788,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0050106036375363e-05,
+      "loss": 3.6776,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0041720088864837e-05,
+      "loss": 3.6755,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0033334141354317e-05,
+      "loss": 3.6766,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024948193843797e-05,
+      "loss": 3.6827,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0016562246333276e-05,
+      "loss": 3.6687,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0008176298822756e-05,
+      "loss": 3.6817,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.870403528213501,
+      "eval_runtime": 308.5135,
+      "eval_samples_per_second": 1236.87,
+      "eval_steps_per_second": 38.653,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9999790351312236e-05,
+      "loss": 3.6671,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9991420782605445e-05,
+      "loss": 3.6626,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998303483509493e-05,
+      "loss": 3.6734,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997464888758441e-05,
+      "loss": 3.6786,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996626294007389e-05,
+      "loss": 3.6858,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99578933713671e-05,
+      "loss": 3.6667,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994950742385658e-05,
+      "loss": 3.6776,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994112147634606e-05,
+      "loss": 3.664,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993273552883554e-05,
+      "loss": 3.6707,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992436596012875e-05,
+      "loss": 3.6712,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991598001261823e-05,
+      "loss": 3.6706,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.990759406510771e-05,
+      "loss": 3.6732,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989922449640092e-05,
+      "loss": 3.6735,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98908385488904e-05,
+      "loss": 3.6629,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9882452601379886e-05,
+      "loss": 3.665,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9874066653869366e-05,
+      "loss": 3.6553,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9865680706358846e-05,
+      "loss": 3.665,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9857294758848326e-05,
+      "loss": 3.6767,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9848925190141535e-05,
+      "loss": 3.6638,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9840539242631015e-05,
+      "loss": 3.6803,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9832153295120495e-05,
+      "loss": 3.6819,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9823767347609975e-05,
+      "loss": 3.6796,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9815381400099455e-05,
+      "loss": 3.6764,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9806995452588935e-05,
+      "loss": 3.6667,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9798609505078415e-05,
+      "loss": 3.6671,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9790223557567895e-05,
+      "loss": 3.6758,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781853988861104e-05,
+      "loss": 3.6728,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9773468041350584e-05,
+      "loss": 3.6662,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976508209384007e-05,
+      "loss": 3.6662,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.975669614632955e-05,
+      "loss": 3.6653,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974832657762276e-05,
+      "loss": 3.66,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973994063011224e-05,
+      "loss": 3.6741,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973155468260172e-05,
+      "loss": 3.6729,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.97231687350912e-05,
+      "loss": 3.6735,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971479916638441e-05,
+      "loss": 3.6633,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970641321887389e-05,
+      "loss": 3.6724,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969802727136337e-05,
+      "loss": 3.6691,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968964132385284e-05,
+      "loss": 3.6518,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968127175514606e-05,
+      "loss": 3.6606,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967288580763554e-05,
+      "loss": 3.6544,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9664499860125024e-05,
+      "loss": 3.6634,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96561139126145e-05,
+      "loss": 3.673,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964774434390771e-05,
+      "loss": 3.6663,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963935839639719e-05,
+      "loss": 3.6596,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963097244888667e-05,
+      "loss": 3.6718,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9622586501376146e-05,
+      "loss": 3.6632,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961421693266936e-05,
+      "loss": 3.6751,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960583098515884e-05,
+      "loss": 3.6652,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9597445037648315e-05,
+      "loss": 3.6403,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958907546894153e-05,
+      "loss": 3.6792,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958068952143101e-05,
+      "loss": 3.6571,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957230357392049e-05,
+      "loss": 3.6667,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956391762640997e-05,
+      "loss": 3.6595,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955554805770319e-05,
+      "loss": 3.6585,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954716211019267e-05,
+      "loss": 3.6503,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953877616268215e-05,
+      "loss": 3.6593,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953039021517162e-05,
+      "loss": 3.6767,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9522020646464836e-05,
+      "loss": 3.6497,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9513634698954316e-05,
+      "loss": 3.6715,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950524875144379e-05,
+      "loss": 3.6627,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949686280393327e-05,
+      "loss": 3.6447,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948847685642275e-05,
+      "loss": 3.6741,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948009090891223e-05,
+      "loss": 3.6522,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947170496140171e-05,
+      "loss": 3.6408,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9463319013891195e-05,
+      "loss": 3.6573,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9454949445184404e-05,
+      "loss": 3.6653,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9446563497673884e-05,
+      "loss": 3.6492,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9438177550163364e-05,
+      "loss": 3.6597,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9429791602652844e-05,
+      "loss": 3.6456,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9421422033946053e-05,
+      "loss": 3.65,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9413036086435533e-05,
+      "loss": 3.6625,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940465013892501e-05,
+      "loss": 3.6586,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939626419141449e-05,
+      "loss": 3.6638,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93878946227077e-05,
+      "loss": 3.6584,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937950867519718e-05,
+      "loss": 3.6707,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937112272768666e-05,
+      "loss": 3.6521,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936273678017615e-05,
+      "loss": 3.6631,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935436721146936e-05,
+      "loss": 3.6509,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.934598126395884e-05,
+      "loss": 3.6474,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933761169525205e-05,
+      "loss": 3.661,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932922574774153e-05,
+      "loss": 3.654,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932083980023101e-05,
+      "loss": 3.6652,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931245385272049e-05,
+      "loss": 3.6689,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930406790520997e-05,
+      "loss": 3.641,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929568195769945e-05,
+      "loss": 3.6435,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928729601018893e-05,
+      "loss": 3.6576,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9278926441482136e-05,
+      "loss": 3.6613,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9270540493971616e-05,
+      "loss": 3.6587,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92621545464611e-05,
+      "loss": 3.6494,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925376859895058e-05,
+      "loss": 3.6594,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924538265144006e-05,
+      "loss": 3.6533,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923699670392954e-05,
+      "loss": 3.651,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922861075641902e-05,
+      "loss": 3.6497,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92202248089085e-05,
+      "loss": 3.6588,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921187161900544e-05,
+      "loss": 3.6648,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920348567149492e-05,
+      "loss": 3.6611,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91950997239844e-05,
+      "loss": 3.6443,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918671377647388e-05,
+      "loss": 3.653,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917832782896336e-05,
+      "loss": 3.6473,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916994188145284e-05,
+      "loss": 3.6545,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916155593394232e-05,
+      "loss": 3.6553,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91531699864318e-05,
+      "loss": 3.6484,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9144800417725016e-05,
+      "loss": 3.6524,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9136414470214496e-05,
+      "loss": 3.6432,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9128028522703976e-05,
+      "loss": 3.6496,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911964257519345e-05,
+      "loss": 3.6399,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9111273006486665e-05,
+      "loss": 3.6513,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9102887058976145e-05,
+      "loss": 3.6501,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9094501111465625e-05,
+      "loss": 3.6566,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90861151639551e-05,
+      "loss": 3.6438,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9077745595248314e-05,
+      "loss": 3.6438,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9069359647737794e-05,
+      "loss": 3.6474,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9060973700227274e-05,
+      "loss": 3.657,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9052587752716754e-05,
+      "loss": 3.6516,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904421818400997e-05,
+      "loss": 3.653,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903583223649945e-05,
+      "loss": 3.6545,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902744628898892e-05,
+      "loss": 3.6652,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90190603414784e-05,
+      "loss": 3.651,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901069077277162e-05,
+      "loss": 3.6522,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90023048252611e-05,
+      "loss": 3.6501,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899391887775057e-05,
+      "loss": 3.6567,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898553293024005e-05,
+      "loss": 3.6502,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897716336153327e-05,
+      "loss": 3.6446,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896877741402275e-05,
+      "loss": 3.6506,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8960407845315964e-05,
+      "loss": 3.6554,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8952021897805444e-05,
+      "loss": 3.6464,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8943635950294924e-05,
+      "loss": 3.637,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89352500027844e-05,
+      "loss": 3.6426,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892686405527388e-05,
+      "loss": 3.6408,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891849448656709e-05,
+      "loss": 3.6522,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891010853905657e-05,
+      "loss": 3.6553,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8901722591546046e-05,
+      "loss": 3.662,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8893336644035526e-05,
+      "loss": 3.6354,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8884950696525006e-05,
+      "loss": 3.6591,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8876564749014486e-05,
+      "loss": 3.6515,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8868178801503966e-05,
+      "loss": 3.6493,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8859792853993446e-05,
+      "loss": 3.6431,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885142328528666e-05,
+      "loss": 3.6481,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884303733777614e-05,
+      "loss": 3.6446,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883465139026562e-05,
+      "loss": 3.6625,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88262654427551e-05,
+      "loss": 3.6395,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881789587404831e-05,
+      "loss": 3.6525,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880950992653779e-05,
+      "loss": 3.6545,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8801140357831e-05,
+      "loss": 3.6539,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879275441032048e-05,
+      "loss": 3.6431,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878436846280996e-05,
+      "loss": 3.6533,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877598251529944e-05,
+      "loss": 3.6545,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8767612946592655e-05,
+      "loss": 3.6426,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8759226999082135e-05,
+      "loss": 3.6543,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.865609884262085,
+      "eval_runtime": 303.5349,
+      "eval_samples_per_second": 1257.157,
+      "eval_steps_per_second": 39.287,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750841051571615e-05,
+      "loss": 3.6435,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8742455104061095e-05,
+      "loss": 3.6321,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8734069156550575e-05,
+      "loss": 3.6458,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8725683209040055e-05,
+      "loss": 3.6545,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8717297261529535e-05,
+      "loss": 3.6579,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8708911314019015e-05,
+      "loss": 3.6422,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8700525366508495e-05,
+      "loss": 3.6507,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8692139418997975e-05,
+      "loss": 3.6343,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8683769850291184e-05,
+      "loss": 3.6456,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8675383902780664e-05,
+      "loss": 3.6445,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8666997955270144e-05,
+      "loss": 3.6441,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8658612007759624e-05,
+      "loss": 3.6486,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865025881785657e-05,
+      "loss": 3.6534,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864187287034605e-05,
+      "loss": 3.6349,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863348692283553e-05,
+      "loss": 3.6374,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.862510097532501e-05,
+      "loss": 3.6239,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861673140661822e-05,
+      "loss": 3.644,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86083454591077e-05,
+      "loss": 3.6505,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859995951159718e-05,
+      "loss": 3.6373,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859157356408666e-05,
+      "loss": 3.6529,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858318761657614e-05,
+      "loss": 3.6578,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857480166906562e-05,
+      "loss": 3.6514,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85664157215551e-05,
+      "loss": 3.6481,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8558046152848307e-05,
+      "loss": 3.6421,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8549660205337787e-05,
+      "loss": 3.6453,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854127425782727e-05,
+      "loss": 3.6458,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853288831031675e-05,
+      "loss": 3.6489,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852450236280623e-05,
+      "loss": 3.6474,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8516116415295706e-05,
+      "loss": 3.6368,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8507730467785186e-05,
+      "loss": 3.6352,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84993608990784e-05,
+      "loss": 3.6366,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849097495156788e-05,
+      "loss": 3.6489,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848260538286109e-05,
+      "loss": 3.6504,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847421943535057e-05,
+      "loss": 3.6479,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846583348784005e-05,
+      "loss": 3.6382,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8457447540329524e-05,
+      "loss": 3.6402,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844906159281901e-05,
+      "loss": 3.6471,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844067564530849e-05,
+      "loss": 3.631,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843228969779797e-05,
+      "loss": 3.6333,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842390375028745e-05,
+      "loss": 3.6337,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841553418158066e-05,
+      "loss": 3.6365,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840714823407014e-05,
+      "loss": 3.6462,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839876228655962e-05,
+      "loss": 3.6431,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83903763390491e-05,
+      "loss": 3.6343,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8382023149146045e-05,
+      "loss": 3.6464,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8373637201635525e-05,
+      "loss": 3.6362,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8365251254125e-05,
+      "loss": 3.6459,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835686530661448e-05,
+      "loss": 3.6469,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8348495737907694e-05,
+      "loss": 3.6151,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834010979039718e-05,
+      "loss": 3.6481,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8331723842886654e-05,
+      "loss": 3.634,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8323337895376134e-05,
+      "loss": 3.6406,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314951947865614e-05,
+      "loss": 3.6363,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830658237915883e-05,
+      "loss": 3.6299,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82981964316483e-05,
+      "loss": 3.6284,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828981048413778e-05,
+      "loss": 3.6305,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828142453662726e-05,
+      "loss": 3.6513,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827305496792047e-05,
+      "loss": 3.6248,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826466902040995e-05,
+      "loss": 3.6473,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825628307289943e-05,
+      "loss": 3.6376,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824789712538892e-05,
+      "loss": 3.621,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82395111778784e-05,
+      "loss": 3.6456,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823114160917161e-05,
+      "loss": 3.6264,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822275566166109e-05,
+      "loss": 3.6176,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821436971415057e-05,
+      "loss": 3.6278,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820598376664005e-05,
+      "loss": 3.643,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819759781912953e-05,
+      "loss": 3.6209,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818921187161901e-05,
+      "loss": 3.6392,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818082592410849e-05,
+      "loss": 3.6226,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817243997659797e-05,
+      "loss": 3.6232,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8164070407891176e-05,
+      "loss": 3.6359,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8155684460380656e-05,
+      "loss": 3.6348,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8147298512870136e-05,
+      "loss": 3.639,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8138912565359616e-05,
+      "loss": 3.6367,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813054299665283e-05,
+      "loss": 3.6421,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812215704914231e-05,
+      "loss": 3.6321,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811377110163179e-05,
+      "loss": 3.6377,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810538515412127e-05,
+      "loss": 3.6219,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809701558541448e-05,
+      "loss": 3.6262,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808864601670769e-05,
+      "loss": 3.6353,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808026006919717e-05,
+      "loss": 3.6253,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807187412168665e-05,
+      "loss": 3.6407,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806348817417613e-05,
+      "loss": 3.6418,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805510222666561e-05,
+      "loss": 3.6233,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804671627915509e-05,
+      "loss": 3.6197,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803833033164457e-05,
+      "loss": 3.6329,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8029960762937786e-05,
+      "loss": 3.6374,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8021574815427266e-05,
+      "loss": 3.6356,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8013188867916746e-05,
+      "loss": 3.6229,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8004802920406225e-05,
+      "loss": 3.6367,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7996416972895705e-05,
+      "loss": 3.6266,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7988047404188915e-05,
+      "loss": 3.6269,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7979661456678394e-05,
+      "loss": 3.625,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7971275509167874e-05,
+      "loss": 3.6315,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7962889561657354e-05,
+      "loss": 3.6394,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7954503614146834e-05,
+      "loss": 3.6379,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794611766663631e-05,
+      "loss": 3.618,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793773171912579e-05,
+      "loss": 3.63,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7929345771615274e-05,
+      "loss": 3.62,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792097620290849e-05,
+      "loss": 3.6342,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.791259025539796e-05,
+      "loss": 3.6311,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790420430788744e-05,
+      "loss": 3.6267,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789581836037692e-05,
+      "loss": 3.6296,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788744879167013e-05,
+      "loss": 3.6194,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787906284415961e-05,
+      "loss": 3.6247,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787069327545283e-05,
+      "loss": 3.6186,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786230732794231e-05,
+      "loss": 3.6264,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785392138043178e-05,
+      "loss": 3.628,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784553543292126e-05,
+      "loss": 3.6286,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783714948541074e-05,
+      "loss": 3.6204,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782876353790023e-05,
+      "loss": 3.6197,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782039396919344e-05,
+      "loss": 3.6226,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781200802168292e-05,
+      "loss": 3.6345,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78036220741724e-05,
+      "loss": 3.6295,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779523612666188e-05,
+      "loss": 3.6281,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778685017915136e-05,
+      "loss": 3.6303,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777846423164084e-05,
+      "loss": 3.6402,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7770094662934046e-05,
+      "loss": 3.627,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7761708715423526e-05,
+      "loss": 3.63,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7753322767913006e-05,
+      "loss": 3.6265,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744936820402486e-05,
+      "loss": 3.6326,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7736550872891966e-05,
+      "loss": 3.6307,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7728164925381446e-05,
+      "loss": 3.6216,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7719778977870926e-05,
+      "loss": 3.6217,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771140940916414e-05,
+      "loss": 3.6318,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770302346165362e-05,
+      "loss": 3.6204,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76946375141431e-05,
+      "loss": 3.6161,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768625156663258e-05,
+      "loss": 3.6175,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767786561912206e-05,
+      "loss": 3.6146,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766949605041527e-05,
+      "loss": 3.6344,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766111010290475e-05,
+      "loss": 3.6258,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765272415539423e-05,
+      "loss": 3.6387,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764433820788371e-05,
+      "loss": 3.6156,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763595226037319e-05,
+      "loss": 3.6316,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762756631286267e-05,
+      "loss": 3.626,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761918036535215e-05,
+      "loss": 3.627,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761079441784163e-05,
+      "loss": 3.6208,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7602424849134846e-05,
+      "loss": 3.6226,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7594038901624326e-05,
+      "loss": 3.6199,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75856529541138e-05,
+      "loss": 3.6415,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757726700660328e-05,
+      "loss": 3.6156,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7568897437896495e-05,
+      "loss": 3.6282,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756051149038597e-05,
+      "loss": 3.6339,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7552141921679184e-05,
+      "loss": 3.6256,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7543755974168664e-05,
+      "loss": 3.6206,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7535370026658144e-05,
+      "loss": 3.6293,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752698407914762e-05,
+      "loss": 3.6311,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751861451044083e-05,
+      "loss": 3.6236,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751022856293032e-05,
+      "loss": 3.6302,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.864367723464966,
+      "eval_runtime": 308.893,
+      "eval_samples_per_second": 1235.35,
+      "eval_steps_per_second": 38.606,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750184261541979e-05,
+      "loss": 3.6164,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749345666790927e-05,
+      "loss": 3.6101,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748507072039875e-05,
+      "loss": 3.6235,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747668477288823e-05,
+      "loss": 3.6305,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.746831520418144e-05,
+      "loss": 3.6355,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745992925667092e-05,
+      "loss": 3.6186,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74515433091604e-05,
+      "loss": 3.627,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744315736164988e-05,
+      "loss": 3.6161,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743477141413936e-05,
+      "loss": 3.6207,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742638546662884e-05,
+      "loss": 3.6219,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741799951911832e-05,
+      "loss": 3.6192,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740962995041154e-05,
+      "loss": 3.6268,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7401260381704747e-05,
+      "loss": 3.6277,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7392874434194226e-05,
+      "loss": 3.6119,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7384488486683706e-05,
+      "loss": 3.6154,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7376102539173186e-05,
+      "loss": 3.6031,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7367732970466395e-05,
+      "loss": 3.6238,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7359347022955875e-05,
+      "loss": 3.6246,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7350961075445355e-05,
+      "loss": 3.6135,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7342575127934835e-05,
+      "loss": 3.6236,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7334189180424315e-05,
+      "loss": 3.6398,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7325819611717524e-05,
+      "loss": 3.6298,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7317433664207004e-05,
+      "loss": 3.6247,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730904771669649e-05,
+      "loss": 3.6182,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730066176918597e-05,
+      "loss": 3.625,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729227582167545e-05,
+      "loss": 3.6203,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728388987416493e-05,
+      "loss": 3.6202,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727550392665441e-05,
+      "loss": 3.6252,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726711797914389e-05,
+      "loss": 3.6154,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72587484104371e-05,
+      "loss": 3.6136,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.725036246292658e-05,
+      "loss": 3.6143,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724197651541606e-05,
+      "loss": 3.6255,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723359056790554e-05,
+      "loss": 3.6242,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722522099919875e-05,
+      "loss": 3.6281,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721683505168823e-05,
+      "loss": 3.619,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720844910417771e-05,
+      "loss": 3.6173,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720006315666719e-05,
+      "loss": 3.6237,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7191677209156676e-05,
+      "loss": 3.6109,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7183291261646155e-05,
+      "loss": 3.6113,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7174921692939365e-05,
+      "loss": 3.6101,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7166535745428845e-05,
+      "loss": 3.6106,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7158149797918324e-05,
+      "loss": 3.6208,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7149763850407804e-05,
+      "loss": 3.6194,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714137790289728e-05,
+      "loss": 3.6176,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7133008334190493e-05,
+      "loss": 3.6173,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7124622386679973e-05,
+      "loss": 3.6171,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7116236439169453e-05,
+      "loss": 3.6244,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7107850491658927e-05,
+      "loss": 3.6235,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709948092295214e-05,
+      "loss": 3.5919,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709109497544163e-05,
+      "loss": 3.6254,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70827090279311e-05,
+      "loss": 3.6113,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707432308042058e-05,
+      "loss": 3.6191,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706593713291006e-05,
+      "loss": 3.6172,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705756756420328e-05,
+      "loss": 3.6038,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704918161669275e-05,
+      "loss": 3.6087,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704079566918223e-05,
+      "loss": 3.6046,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703240972167171e-05,
+      "loss": 3.6253,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702404015296493e-05,
+      "loss": 3.6068,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70156542054544e-05,
+      "loss": 3.6213,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700726825794388e-05,
+      "loss": 3.6171,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699888231043337e-05,
+      "loss": 3.6037,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699049636292285e-05,
+      "loss": 3.6208,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6982126794216056e-05,
+      "loss": 3.6029,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6973740846705536e-05,
+      "loss": 3.5991,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6965354899195016e-05,
+      "loss": 3.603,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6956968951684496e-05,
+      "loss": 3.6206,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6948583004173976e-05,
+      "loss": 3.5985,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6940197056663456e-05,
+      "loss": 3.6193,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6931811109152936e-05,
+      "loss": 3.5957,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6923441540446145e-05,
+      "loss": 3.6038,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6915055592935625e-05,
+      "loss": 3.6129,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6906669645425105e-05,
+      "loss": 3.6122,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6898283697914585e-05,
+      "loss": 3.6168,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6889897750404065e-05,
+      "loss": 3.6163,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6881511802893545e-05,
+      "loss": 3.622,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687314223418676e-05,
+      "loss": 3.6048,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686475628667624e-05,
+      "loss": 3.6164,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685637033916572e-05,
+      "loss": 3.6023,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68479843916552e-05,
+      "loss": 3.6057,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683961482294841e-05,
+      "loss": 3.6099,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683122887543789e-05,
+      "loss": 3.6037,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682284292792737e-05,
+      "loss": 3.6185,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681445698041685e-05,
+      "loss": 3.6208,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680607103290633e-05,
+      "loss": 3.6079,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679768508539581e-05,
+      "loss": 3.5912,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678931551668902e-05,
+      "loss": 3.613,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.67809295691785e-05,
+      "loss": 3.6123,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6772543621667985e-05,
+      "loss": 3.6199,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6764157674157465e-05,
+      "loss": 3.5967,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675577172664694e-05,
+      "loss": 3.6148,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674738577913642e-05,
+      "loss": 3.6068,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6739016210429634e-05,
+      "loss": 3.6037,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6730630262919114e-05,
+      "loss": 3.6019,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672224431540859e-05,
+      "loss": 3.6115,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671385836789807e-05,
+      "loss": 3.6136,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670548879919128e-05,
+      "loss": 3.6177,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6697102851680756e-05,
+      "loss": 3.5987,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6688716904170236e-05,
+      "loss": 3.6073,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668033095665972e-05,
+      "loss": 3.6007,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66719450091492e-05,
+      "loss": 3.6112,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666357544044241e-05,
+      "loss": 3.6074,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665518949293189e-05,
+      "loss": 3.6056,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664680354542137e-05,
+      "loss": 3.6038,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663841759791085e-05,
+      "loss": 3.5973,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663003165040033e-05,
+      "loss": 3.6074,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662166208169354e-05,
+      "loss": 3.5917,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661327613418302e-05,
+      "loss": 3.6029,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66048901866725e-05,
+      "loss": 3.6036,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659650423916198e-05,
+      "loss": 3.6132,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658811829165146e-05,
+      "loss": 3.597,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657973234414094e-05,
+      "loss": 3.601,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6571362775434156e-05,
+      "loss": 3.6008,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6562976827923636e-05,
+      "loss": 3.6132,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6554590880413116e-05,
+      "loss": 3.6087,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6546204932902596e-05,
+      "loss": 3.6068,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6537835364195805e-05,
+      "loss": 3.6111,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6529449416685285e-05,
+      "loss": 3.617,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6521063469174765e-05,
+      "loss": 3.605,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6512677521664245e-05,
+      "loss": 3.609,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6504291574153725e-05,
+      "loss": 3.6057,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6495922005446934e-05,
+      "loss": 3.6113,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6487536057936414e-05,
+      "loss": 3.6058,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6479150110425894e-05,
+      "loss": 3.6003,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6470764162915374e-05,
+      "loss": 3.6012,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646239459420859e-05,
+      "loss": 3.6108,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645400864669807e-05,
+      "loss": 3.5979,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644562269918755e-05,
+      "loss": 3.5996,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643723675167703e-05,
+      "loss": 3.5896,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642886718297024e-05,
+      "loss": 3.5976,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642048123545972e-05,
+      "loss": 3.6099,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64120952879492e-05,
+      "loss": 3.6081,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640370934043868e-05,
+      "loss": 3.6212,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639532339292816e-05,
+      "loss": 3.5925,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638693744541764e-05,
+      "loss": 3.6111,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637855149790712e-05,
+      "loss": 3.6034,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637018192920033e-05,
+      "loss": 3.6068,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6361795981689815e-05,
+      "loss": 3.5976,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6353410034179295e-05,
+      "loss": 3.5999,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6345024086668775e-05,
+      "loss": 3.6031,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6336654517961984e-05,
+      "loss": 3.6196,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6328268570451464e-05,
+      "loss": 3.5925,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6319882622940944e-05,
+      "loss": 3.6084,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631149667543042e-05,
+      "loss": 3.6097,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630312710672363e-05,
+      "loss": 3.6045,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629474115921311e-05,
+      "loss": 3.6011,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628635521170259e-05,
+      "loss": 3.6092,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6277969264192066e-05,
+      "loss": 3.6064,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626959969548528e-05,
+      "loss": 3.6053,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626121374797477e-05,
+      "loss": 3.6039,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.861919403076172,
+      "eval_runtime": 301.9684,
+      "eval_samples_per_second": 1263.679,
+      "eval_steps_per_second": 39.491,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625282780046425e-05,
+      "loss": 3.5901,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624444185295372e-05,
+      "loss": 3.5908,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62360559054432e-05,
+      "loss": 3.6035,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622766995793268e-05,
+      "loss": 3.6075,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621930038922589e-05,
+      "loss": 3.6155,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621091444171537e-05,
+      "loss": 3.5981,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620252849420485e-05,
+      "loss": 3.6088,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619414254669433e-05,
+      "loss": 3.5954,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618577297798754e-05,
+      "loss": 3.5972,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617738703047702e-05,
+      "loss": 3.5974,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6169001082966506e-05,
+      "loss": 3.6019,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160615135455986e-05,
+      "loss": 3.6033,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6152245566749195e-05,
+      "loss": 3.6087,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6143859619238675e-05,
+      "loss": 3.592,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6135473671728155e-05,
+      "loss": 3.593,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6127087724217635e-05,
+      "loss": 3.5814,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6118701776707115e-05,
+      "loss": 3.6012,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6110315829196595e-05,
+      "loss": 3.6034,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6101929881686075e-05,
+      "loss": 3.5914,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6093543934175555e-05,
+      "loss": 3.6051,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6085157986665035e-05,
+      "loss": 3.6191,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6076788417958244e-05,
+      "loss": 3.6077,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6068402470447724e-05,
+      "loss": 3.6031,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6060016522937204e-05,
+      "loss": 3.5989,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6051630575426684e-05,
+      "loss": 3.6044,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604324462791617e-05,
+      "loss": 3.6032,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603485868040565e-05,
+      "loss": 3.5992,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602647273289513e-05,
+      "loss": 3.601,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601810316418834e-05,
+      "loss": 3.5972,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600973359548155e-05,
+      "loss": 3.5917,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600134764797103e-05,
+      "loss": 3.5996,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599296170046051e-05,
+      "loss": 3.598,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598457575294999e-05,
+      "loss": 3.6041,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597618980543947e-05,
+      "loss": 3.6079,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596782023673268e-05,
+      "loss": 3.5977,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595943428922216e-05,
+      "loss": 3.5955,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595104834171164e-05,
+      "loss": 3.6052,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5942662394201124e-05,
+      "loss": 3.5894,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5934276446690604e-05,
+      "loss": 3.5926,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592590687798381e-05,
+      "loss": 3.5877,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591752093047329e-05,
+      "loss": 3.5947,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590913498296277e-05,
+      "loss": 3.5954,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590074903545225e-05,
+      "loss": 3.599,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589237946674546e-05,
+      "loss": 3.6006,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588399351923494e-05,
+      "loss": 3.5956,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587560757172442e-05,
+      "loss": 3.601,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58672216242139e-05,
+      "loss": 3.6017,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585885205550711e-05,
+      "loss": 3.605,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585046610799659e-05,
+      "loss": 3.5737,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584208016048608e-05,
+      "loss": 3.6053,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583369421297556e-05,
+      "loss": 3.5885,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582530826546503e-05,
+      "loss": 3.5976,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581692231795451e-05,
+      "loss": 3.5941,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580855274924773e-05,
+      "loss": 3.5884,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58001668017372e-05,
+      "loss": 3.5842,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579178085422668e-05,
+      "loss": 3.5829,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578339490671616e-05,
+      "loss": 3.6107,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5775025338009376e-05,
+      "loss": 3.5829,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576663939049885e-05,
+      "loss": 3.6023,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575825344298833e-05,
+      "loss": 3.5937,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5749867495477816e-05,
+      "loss": 3.5843,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5741481547967296e-05,
+      "loss": 3.6008,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5733111979260505e-05,
+      "loss": 3.5863,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5724726031749985e-05,
+      "loss": 3.576,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5716340084239465e-05,
+      "loss": 3.5848,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5707954136728944e-05,
+      "loss": 3.6015,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5699584568022154e-05,
+      "loss": 3.5807,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5691198620511634e-05,
+      "loss": 3.6005,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5682812673001114e-05,
+      "loss": 3.5731,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5674426725490593e-05,
+      "loss": 3.5854,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56660571567838e-05,
+      "loss": 3.5935,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565767120927328e-05,
+      "loss": 3.5935,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564928526176277e-05,
+      "loss": 3.5933,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564089931425225e-05,
+      "loss": 3.5964,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563251336674173e-05,
+      "loss": 3.6008,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562414379803494e-05,
+      "loss": 3.5869,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561575785052442e-05,
+      "loss": 3.5936,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56073719030139e-05,
+      "loss": 3.5856,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559898595550338e-05,
+      "loss": 3.5881,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559060000799286e-05,
+      "loss": 3.5873,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558223043928607e-05,
+      "loss": 3.5869,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557384449177555e-05,
+      "loss": 3.5973,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556545854426503e-05,
+      "loss": 3.5994,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555707259675451e-05,
+      "loss": 3.5897,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554868664924399e-05,
+      "loss": 3.5696,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554030070173347e-05,
+      "loss": 3.5927,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553193113302668e-05,
+      "loss": 3.5946,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552354518551616e-05,
+      "loss": 3.601,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551515923800564e-05,
+      "loss": 3.5779,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550677329049512e-05,
+      "loss": 3.5904,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54983873429846e-05,
+      "loss": 3.5944,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549001777427781e-05,
+      "loss": 3.5812,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548163182676729e-05,
+      "loss": 3.5803,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547324587925677e-05,
+      "loss": 3.5937,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546485993174625e-05,
+      "loss": 3.5917,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545649036303946e-05,
+      "loss": 3.6,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544810441552894e-05,
+      "loss": 3.5819,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543971846801842e-05,
+      "loss": 3.584,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54313325205079e-05,
+      "loss": 3.5805,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542294657299739e-05,
+      "loss": 3.5933,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5414577004290596e-05,
+      "loss": 3.5913,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5406191056780076e-05,
+      "loss": 3.5849,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5397805109269556e-05,
+      "loss": 3.583,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5389419161759036e-05,
+      "loss": 3.5807,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538103321424851e-05,
+      "loss": 3.587,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5372663645541725e-05,
+      "loss": 3.5726,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5364277698031205e-05,
+      "loss": 3.5853,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5355891750520685e-05,
+      "loss": 3.5813,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534750580301016e-05,
+      "loss": 3.5907,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533911985549964e-05,
+      "loss": 3.5812,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5330750286792854e-05,
+      "loss": 3.5888,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5322364339282334e-05,
+      "loss": 3.5774,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5313978391771814e-05,
+      "loss": 3.5948,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5305592444261294e-05,
+      "loss": 3.5861,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529722287555451e-05,
+      "loss": 3.5891,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528883692804398e-05,
+      "loss": 3.587,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528045098053346e-05,
+      "loss": 3.6031,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527206503302294e-05,
+      "loss": 3.5837,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526369546431616e-05,
+      "loss": 3.5895,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525530951680563e-05,
+      "loss": 3.5882,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524692356929511e-05,
+      "loss": 3.5898,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523853762178459e-05,
+      "loss": 3.5848,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523015167427408e-05,
+      "loss": 3.5852,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522176572676356e-05,
+      "loss": 3.581,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521339615805677e-05,
+      "loss": 3.589,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520501021054625e-05,
+      "loss": 3.5766,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519662426303573e-05,
+      "loss": 3.5802,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518823831552521e-05,
+      "loss": 3.5728,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517986874681842e-05,
+      "loss": 3.5836,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51714827993079e-05,
+      "loss": 3.5889,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516309685179738e-05,
+      "loss": 3.5873,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515471090428686e-05,
+      "loss": 3.6011,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5146341335580066e-05,
+      "loss": 3.5767,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5137955388069546e-05,
+      "loss": 3.5866,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512956944055903e-05,
+      "loss": 3.5899,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512118349304851e-05,
+      "loss": 3.5873,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511279754553799e-05,
+      "loss": 3.5789,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510441159802747e-05,
+      "loss": 3.5763,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509602565051695e-05,
+      "loss": 3.5872,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508765608181016e-05,
+      "loss": 3.594,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507927013429964e-05,
+      "loss": 3.5801,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507088418678912e-05,
+      "loss": 3.5885,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50624982392786e-05,
+      "loss": 3.5904,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505412867057181e-05,
+      "loss": 3.5852,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504574272306129e-05,
+      "loss": 3.5843,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503735677555077e-05,
+      "loss": 3.5881,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502897082804025e-05,
+      "loss": 3.587,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5020601259333466e-05,
+      "loss": 3.5846,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5012215311822946e-05,
+      "loss": 3.5885,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.862126350402832,
+      "eval_runtime": 304.3076,
+      "eval_samples_per_second": 1253.965,
+      "eval_steps_per_second": 39.187,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5003829364312426e-05,
+      "loss": 3.5815,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4995443416801906e-05,
+      "loss": 3.5677,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4987057469291386e-05,
+      "loss": 3.5849,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4978687900584595e-05,
+      "loss": 3.5898,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4970301953074075e-05,
+      "loss": 3.5919,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4961916005563555e-05,
+      "loss": 3.5819,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4953530058053035e-05,
+      "loss": 3.5853,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4945144110542515e-05,
+      "loss": 3.5789,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4936774541835724e-05,
+      "loss": 3.5774,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4928388594325204e-05,
+      "loss": 3.5794,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4920002646814684e-05,
+      "loss": 3.5844,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4911616699304164e-05,
+      "loss": 3.582,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4903230751793644e-05,
+      "loss": 3.5919,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489486118308686e-05,
+      "loss": 3.57,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488647523557634e-05,
+      "loss": 3.5752,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487808928806582e-05,
+      "loss": 3.5664,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486970334055529e-05,
+      "loss": 3.5835,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486131739304477e-05,
+      "loss": 3.5812,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.485294782433799e-05,
+      "loss": 3.5798,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.484456187682746e-05,
+      "loss": 3.581,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.483617592931694e-05,
+      "loss": 3.598,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482778998180642e-05,
+      "loss": 3.5879,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48194040342959e-05,
+      "loss": 3.584,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481103446558912e-05,
+      "loss": 3.5774,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48026485180786e-05,
+      "loss": 3.5911,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479426257056808e-05,
+      "loss": 3.5822,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478587662305756e-05,
+      "loss": 3.5779,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4777507054350766e-05,
+      "loss": 3.5882,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4769121106840246e-05,
+      "loss": 3.5704,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4760735159329726e-05,
+      "loss": 3.575,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4752349211819206e-05,
+      "loss": 3.5806,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4743979643112415e-05,
+      "loss": 3.5773,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4735593695601895e-05,
+      "loss": 3.5869,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4727207748091375e-05,
+      "loss": 3.5883,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4718821800580855e-05,
+      "loss": 3.5767,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471045223187407e-05,
+      "loss": 3.5794,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470206628436355e-05,
+      "loss": 3.586,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469368033685303e-05,
+      "loss": 3.5703,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468529438934251e-05,
+      "loss": 3.5741,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467690844183199e-05,
+      "loss": 3.5726,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46685388731252e-05,
+      "loss": 3.5728,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466015292561468e-05,
+      "loss": 3.5747,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465176697810416e-05,
+      "loss": 3.5819,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464338103059364e-05,
+      "loss": 3.5817,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463499508308312e-05,
+      "loss": 3.5768,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462662551437633e-05,
+      "loss": 3.5802,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461823956686581e-05,
+      "loss": 3.5839,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4609853619355296e-05,
+      "loss": 3.5842,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4601467671844776e-05,
+      "loss": 3.5612,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4593098103137985e-05,
+      "loss": 3.5821,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4584712155627465e-05,
+      "loss": 3.57,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4576326208116945e-05,
+      "loss": 3.5762,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4567940260606425e-05,
+      "loss": 3.5787,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4559570691899634e-05,
+      "loss": 3.5688,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4551184744389114e-05,
+      "loss": 3.5667,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4542798796878594e-05,
+      "loss": 3.5621,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4534412849368073e-05,
+      "loss": 3.593,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452604328066128e-05,
+      "loss": 3.5675,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451765733315076e-05,
+      "loss": 3.5816,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450927138564025e-05,
+      "loss": 3.5747,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450088543812973e-05,
+      "loss": 3.5664,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449251586942294e-05,
+      "loss": 3.5782,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448412992191242e-05,
+      "loss": 3.5707,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44757439744019e-05,
+      "loss": 3.5614,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446735802689138e-05,
+      "loss": 3.5594,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445897207938086e-05,
+      "loss": 3.5872,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445060251067407e-05,
+      "loss": 3.5646,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444221656316355e-05,
+      "loss": 3.5751,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443383061565303e-05,
+      "loss": 3.557,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442544466814251e-05,
+      "loss": 3.5703,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4417075099435716e-05,
+      "loss": 3.5693,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44086891519252e-05,
+      "loss": 3.5776,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440030320441468e-05,
+      "loss": 3.5768,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439191725690416e-05,
+      "loss": 3.5762,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438353130939364e-05,
+      "loss": 3.5809,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.437516174068685e-05,
+      "loss": 3.5686,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436677579317633e-05,
+      "loss": 3.5742,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435838984566581e-05,
+      "loss": 3.5666,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435000389815529e-05,
+      "loss": 3.57,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.43416343294485e-05,
+      "loss": 3.5735,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433324838193798e-05,
+      "loss": 3.5648,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.432486243442746e-05,
+      "loss": 3.5812,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.431647648691694e-05,
+      "loss": 3.5819,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430810691821016e-05,
+      "loss": 3.573,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4299720970699637e-05,
+      "loss": 3.5492,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4291335023189117e-05,
+      "loss": 3.57,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4282949075678596e-05,
+      "loss": 3.5782,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427456312816807e-05,
+      "loss": 3.5816,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4266193559461286e-05,
+      "loss": 3.5668,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4257807611950765e-05,
+      "loss": 3.5695,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4249421664440245e-05,
+      "loss": 3.5756,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424103571692972e-05,
+      "loss": 3.5634,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42326497694192e-05,
+      "loss": 3.5634,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4224280200712414e-05,
+      "loss": 3.5738,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4215894253201894e-05,
+      "loss": 3.5715,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4207508305691374e-05,
+      "loss": 3.5824,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4199122358180854e-05,
+      "loss": 3.5674,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4190736410670334e-05,
+      "loss": 3.5651,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4182350463159814e-05,
+      "loss": 3.5619,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.417398089445302e-05,
+      "loss": 3.5751,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41655949469425e-05,
+      "loss": 3.5724,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415720899943198e-05,
+      "loss": 3.5662,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414882305192146e-05,
+      "loss": 3.566,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414045348321467e-05,
+      "loss": 3.5596,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413206753570415e-05,
+      "loss": 3.5723,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.412368158819363e-05,
+      "loss": 3.5505,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.411529564068311e-05,
+      "loss": 3.5644,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410692607197633e-05,
+      "loss": 3.5676,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409854012446581e-05,
+      "loss": 3.5713,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409015417695529e-05,
+      "loss": 3.5589,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408176822944477e-05,
+      "loss": 3.5708,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407338228193425e-05,
+      "loss": 3.5615,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406501271322746e-05,
+      "loss": 3.5731,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405662676571694e-05,
+      "loss": 3.5695,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404824081820642e-05,
+      "loss": 3.57,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40398548706959e-05,
+      "loss": 3.5692,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403146892318538e-05,
+      "loss": 3.5857,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4023099354478586e-05,
+      "loss": 3.5645,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4014713406968066e-05,
+      "loss": 3.5696,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4006327459457546e-05,
+      "loss": 3.5752,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3997941511947026e-05,
+      "loss": 3.5691,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398955556443651e-05,
+      "loss": 3.5674,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398118599572972e-05,
+      "loss": 3.566,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.39728000482192e-05,
+      "loss": 3.564,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396441410070868e-05,
+      "loss": 3.5706,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395602815319816e-05,
+      "loss": 3.5556,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394764220568764e-05,
+      "loss": 3.5676,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393925625817712e-05,
+      "loss": 3.5572,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393088668947033e-05,
+      "loss": 3.5665,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392250074195981e-05,
+      "loss": 3.5625,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391411479444929e-05,
+      "loss": 3.5705,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390572884693877e-05,
+      "loss": 3.5818,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389735927823198e-05,
+      "loss": 3.5606,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3888973330721466e-05,
+      "loss": 3.57,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3880587383210946e-05,
+      "loss": 3.5696,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3872201435700426e-05,
+      "loss": 3.5676,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3863815488189906e-05,
+      "loss": 3.5647,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3855445919483115e-05,
+      "loss": 3.5567,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3847059971972595e-05,
+      "loss": 3.5725,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3838674024462075e-05,
+      "loss": 3.577,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3830288076951555e-05,
+      "loss": 3.5636,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382190212944103e-05,
+      "loss": 3.569,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381351618193051e-05,
+      "loss": 3.573,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3805146613223724e-05,
+      "loss": 3.5652,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3796760665713204e-05,
+      "loss": 3.5646,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3788374718202684e-05,
+      "loss": 3.5742,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3779988770692164e-05,
+      "loss": 3.5665,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377161920198538e-05,
+      "loss": 3.5661,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376323325447485e-05,
+      "loss": 3.5743,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.862633466720581,
+      "eval_runtime": 305.3094,
+      "eval_samples_per_second": 1249.85,
+      "eval_steps_per_second": 39.059,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375484730696433e-05,
+      "loss": 3.5708,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374646135945381e-05,
+      "loss": 3.5473,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373807541194329e-05,
+      "loss": 3.5651,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372968946443277e-05,
+      "loss": 3.5753,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372130351692225e-05,
+      "loss": 3.5738,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371293394821546e-05,
+      "loss": 3.5666,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370454800070494e-05,
+      "loss": 3.5655,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369616205319442e-05,
+      "loss": 3.5638,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36877761056839e-05,
+      "loss": 3.558,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367940653697712e-05,
+      "loss": 3.5652,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36710205894666e-05,
+      "loss": 3.5644,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366263464195608e-05,
+      "loss": 3.5654,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3654265073249287e-05,
+      "loss": 3.5764,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3645879125738766e-05,
+      "loss": 3.5548,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3637493178228246e-05,
+      "loss": 3.5546,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3629107230717726e-05,
+      "loss": 3.5464,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3620721283207206e-05,
+      "loss": 3.5669,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3612335335696686e-05,
+      "loss": 3.5637,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3603949388186166e-05,
+      "loss": 3.5649,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3595563440675646e-05,
+      "loss": 3.5624,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3587193871968855e-05,
+      "loss": 3.5796,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357880792445834e-05,
+      "loss": 3.5702,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357042197694782e-05,
+      "loss": 3.5665,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35620360294373e-05,
+      "loss": 3.5615,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355366646073051e-05,
+      "loss": 3.5771,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354528051321999e-05,
+      "loss": 3.5593,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353689456570947e-05,
+      "loss": 3.5639,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352850861819895e-05,
+      "loss": 3.5723,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352012267068843e-05,
+      "loss": 3.5542,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351175310198164e-05,
+      "loss": 3.5587,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350336715447112e-05,
+      "loss": 3.5612,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.34949812069606e-05,
+      "loss": 3.5592,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348659525945008e-05,
+      "loss": 3.5697,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347820931193956e-05,
+      "loss": 3.5712,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346982336442904e-05,
+      "loss": 3.5626,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3461453795722256e-05,
+      "loss": 3.5616,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3453067848211736e-05,
+      "loss": 3.5728,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3444681900701216e-05,
+      "loss": 3.5464,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343629595319069e-05,
+      "loss": 3.56,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3427926384483905e-05,
+      "loss": 3.5577,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3419540436973385e-05,
+      "loss": 3.5525,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3411154489462864e-05,
+      "loss": 3.5586,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340276854195234e-05,
+      "loss": 3.568,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3394398973245554e-05,
+      "loss": 3.5642,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3386013025735033e-05,
+      "loss": 3.5583,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3377627078224513e-05,
+      "loss": 3.561,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3369241130713993e-05,
+      "loss": 3.5631,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336087156200721e-05,
+      "loss": 3.5689,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335248561449669e-05,
+      "loss": 3.5466,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334409966698616e-05,
+      "loss": 3.5669,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333571371947564e-05,
+      "loss": 3.5576,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332732777196512e-05,
+      "loss": 3.5572,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331895820325834e-05,
+      "loss": 3.5626,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331057225574781e-05,
+      "loss": 3.546,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330218630823729e-05,
+      "loss": 3.5557,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329381673953051e-05,
+      "loss": 3.5434,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328543079201998e-05,
+      "loss": 3.5757,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327704484450947e-05,
+      "loss": 3.5493,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326865889699895e-05,
+      "loss": 3.5671,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326027294948843e-05,
+      "loss": 3.5538,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3251903380781636e-05,
+      "loss": 3.552,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3243517433271116e-05,
+      "loss": 3.5582,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3235131485760596e-05,
+      "loss": 3.5556,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3226745538250076e-05,
+      "loss": 3.5476,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3218359590739556e-05,
+      "loss": 3.5396,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3209990022032765e-05,
+      "loss": 3.5642,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3201604074522245e-05,
+      "loss": 3.5518,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3193218127011725e-05,
+      "loss": 3.5544,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3184832179501205e-05,
+      "loss": 3.543,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3176446231990685e-05,
+      "loss": 3.5471,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31680766632839e-05,
+      "loss": 3.5556,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315969071577338e-05,
+      "loss": 3.5607,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315130476826286e-05,
+      "loss": 3.5569,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314291882075234e-05,
+      "loss": 3.5594,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.313454925204555e-05,
+      "loss": 3.5628,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312616330453503e-05,
+      "loss": 3.5516,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311777735702451e-05,
+      "loss": 3.5571,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310939140951399e-05,
+      "loss": 3.5508,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31010218408072e-05,
+      "loss": 3.554,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309263589329668e-05,
+      "loss": 3.557,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308424994578616e-05,
+      "loss": 3.5479,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307586399827564e-05,
+      "loss": 3.5652,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3067494429568854e-05,
+      "loss": 3.5648,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3059108482058334e-05,
+      "loss": 3.551,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3050722534547814e-05,
+      "loss": 3.5378,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3042336587037294e-05,
+      "loss": 3.5561,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30339670183305e-05,
+      "loss": 3.5593,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302558107081998e-05,
+      "loss": 3.5617,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301719512330946e-05,
+      "loss": 3.5474,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300880917579894e-05,
+      "loss": 3.5536,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.300043960709215e-05,
+      "loss": 3.5595,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299205365958163e-05,
+      "loss": 3.5452,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298366771207111e-05,
+      "loss": 3.5474,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297528176456059e-05,
+      "loss": 3.5561,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296691219585381e-05,
+      "loss": 3.5539,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295852624834329e-05,
+      "loss": 3.5665,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295014030083277e-05,
+      "loss": 3.5551,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294175435332225e-05,
+      "loss": 3.5436,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293338478461546e-05,
+      "loss": 3.5484,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292499883710494e-05,
+      "loss": 3.5548,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291661288959442e-05,
+      "loss": 3.5606,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29082269420839e-05,
+      "loss": 3.552,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289984099457338e-05,
+      "loss": 3.5465,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2891471425866586e-05,
+      "loss": 3.5409,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2883085478356066e-05,
+      "loss": 3.5591,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2874699530845546e-05,
+      "loss": 3.5342,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2866313583335026e-05,
+      "loss": 3.5452,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285794401462824e-05,
+      "loss": 3.5536,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284955806711772e-05,
+      "loss": 3.5555,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28411721196072e-05,
+      "loss": 3.5414,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283278617209668e-05,
+      "loss": 3.5539,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282441660338989e-05,
+      "loss": 3.5458,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281603065587937e-05,
+      "loss": 3.555,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280764470836885e-05,
+      "loss": 3.5552,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279925876085833e-05,
+      "loss": 3.5492,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279088919215154e-05,
+      "loss": 3.5537,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278250324464102e-05,
+      "loss": 3.5679,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27741172971305e-05,
+      "loss": 3.5487,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276573134961998e-05,
+      "loss": 3.5556,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2757361780913195e-05,
+      "loss": 3.554,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2748975833402675e-05,
+      "loss": 3.5506,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2740589885892155e-05,
+      "loss": 3.5517,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2732203938381635e-05,
+      "loss": 3.548,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2723834369674844e-05,
+      "loss": 3.5476,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2715448422164324e-05,
+      "loss": 3.5541,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2707062474653804e-05,
+      "loss": 3.5437,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2698676527143284e-05,
+      "loss": 3.5485,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2690290579632764e-05,
+      "loss": 3.5418,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268192101092597e-05,
+      "loss": 3.5492,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267353506341545e-05,
+      "loss": 3.5425,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266514911590493e-05,
+      "loss": 3.5602,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265676316839441e-05,
+      "loss": 3.564,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264839359968763e-05,
+      "loss": 3.5407,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264000765217711e-05,
+      "loss": 3.5517,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263162170466659e-05,
+      "loss": 3.5553,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262323575715606e-05,
+      "loss": 3.554,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261486618844928e-05,
+      "loss": 3.5417,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260648024093876e-05,
+      "loss": 3.5439,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259809429342824e-05,
+      "loss": 3.5579,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258970834591771e-05,
+      "loss": 3.5595,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258133877721093e-05,
+      "loss": 3.5429,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257295282970041e-05,
+      "loss": 3.5544,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256456688218989e-05,
+      "loss": 3.5584,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255618093467937e-05,
+      "loss": 3.5511,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254781136597258e-05,
+      "loss": 3.5456,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253942541846206e-05,
+      "loss": 3.5565,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2531039470951536e-05,
+      "loss": 3.55,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2522653523441016e-05,
+      "loss": 3.5554,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251428395473423e-05,
+      "loss": 3.5508,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8622679710388184,
+      "eval_runtime": 302.265,
+      "eval_samples_per_second": 1262.438,
+      "eval_steps_per_second": 39.452,
+      "step": 1068480
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 7.207234546152776e+17,
+  "trial_name": null,
+  "trial_params": null
+}