diff --git "a/checkpoint-457920/trainer_state.json" "b/checkpoint-457920/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457920/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 4.161040306091309,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/lstm/4/checkpoints/checkpoint-457920",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 457920,
+  "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": 10.8203,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5554,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0617,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9928,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9463,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8938,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7206,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6102,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5164,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.447,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3857,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.328,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2684,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.2061,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1519,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0943,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0496,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.0038,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9633,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9267,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8922,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8559,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8247,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.796,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7777,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7387,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7171,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6922,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6726,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6463,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974845433229186e-05,
+      "loss": 5.6222,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974006838478134e-05,
+      "loss": 5.6155,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973168243727082e-05,
+      "loss": 5.5884,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97232964897603e-05,
+      "loss": 5.5783,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5648,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5395,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5286,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4887,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4899,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4549,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4551,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4362,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.435,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4091,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.4015,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3932,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3755,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3797,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3497,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3351,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3448,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3373,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3124,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.301,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2988,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.283,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2966,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2574,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2572,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2491,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2377,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2392,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2173,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2118,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2034,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2166,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1889,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1878,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1617,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1689,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1778,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1681,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1465,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1595,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1548,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1311,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1289,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1113,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0985,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.0977,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1003,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0981,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.1038,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0892,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0694,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0711,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0705,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0586,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0501,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0399,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0393,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0403,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0274,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0294,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0181,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0141,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9998,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9976,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9913,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.9944,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9953,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9728,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.978,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9667,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9644,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9573,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9506,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9497,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9416,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.939,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9309,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.9213,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.9316,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052682749778396e-05,
+      "loss": 4.9214,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044296802267876e-05,
+      "loss": 4.9222,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9173,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9176,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9086,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.8933,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.901,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.9055,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.8914,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.8845,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.882,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896047008477014e-05,
+      "loss": 4.8806,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8705,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8736,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8628,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8634,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8545,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8728,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.855,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.846,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8419,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8434,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8446,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8441,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8329,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.8338,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8293,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8271,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8143,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8265,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8126,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879281664977466e-05,
+      "loss": 4.7996,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8145,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8026,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8028,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8086,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7898,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.756908416748047,
+      "eval_runtime": 298.6182,
+      "eval_samples_per_second": 1277.856,
+      "eval_steps_per_second": 39.934,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7827,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7867,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.788,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7828,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7875,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7709,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7771,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.7643,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.7734,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7597,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7591,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7594,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7523,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7478,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7445,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7309,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7425,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7357,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.728,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.7474,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.7299,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.7319,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7253,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7435,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.7118,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.7135,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.7168,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851617865474989e-05,
+      "loss": 4.7186,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850779270723937e-05,
+      "loss": 4.703,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7032,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7127,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7096,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.7108,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7054,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.6966,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.7052,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.6758,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.6963,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.6715,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.6844,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840721047352432e-05,
+      "loss": 4.6757,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83988245260138e-05,
+      "loss": 4.6873,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.6754,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838205263099276e-05,
+      "loss": 4.6738,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373666683482236e-05,
+      "loss": 4.676,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365280735971716e-05,
+      "loss": 4.6671,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.682,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6564,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.6505,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833175332473337e-05,
+      "loss": 4.67,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832336737722285e-05,
+      "loss": 4.6691,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831498142971233e-05,
+      "loss": 4.6536,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830659548220181e-05,
+      "loss": 4.6484,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829822591349502e-05,
+      "loss": 4.6545,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82898399659845e-05,
+      "loss": 4.6417,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828145401847398e-05,
+      "loss": 4.6647,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827306807096346e-05,
+      "loss": 4.6385,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826468212345294e-05,
+      "loss": 4.646,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825629617594242e-05,
+      "loss": 4.6395,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82479102284319e-05,
+      "loss": 4.6333,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823952428092138e-05,
+      "loss": 4.6439,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823113833341086e-05,
+      "loss": 4.6299,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822275238590034e-05,
+      "loss": 4.6259,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214382817193557e-05,
+      "loss": 4.6282,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8205996869683037e-05,
+      "loss": 4.6424,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.6241,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189241353465726e-05,
+      "loss": 4.6216,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180855405955206e-05,
+      "loss": 4.6148,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172469458444685e-05,
+      "loss": 4.6161,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164083510934165e-05,
+      "loss": 4.6353,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155697563423645e-05,
+      "loss": 4.632,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147327994716854e-05,
+      "loss": 4.6147,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.627,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.6293,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.6164,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811380058347851e-05,
+      "loss": 4.617,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810541463596799e-05,
+      "loss": 4.6064,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809702868845747e-05,
+      "loss": 4.5952,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808864274094695e-05,
+      "loss": 4.5979,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808025679343642e-05,
+      "loss": 4.6045,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807188722472964e-05,
+      "loss": 4.6056,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806350127721912e-05,
+      "loss": 4.6232,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80551153297086e-05,
+      "loss": 4.6089,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804672938219807e-05,
+      "loss": 4.5892,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803834343468755e-05,
+      "loss": 4.5962,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802995748717703e-05,
+      "loss": 4.6017,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802157153966652e-05,
+      "loss": 4.5964,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013185592156e-05,
+      "loss": 4.5884,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800479964464548e-05,
+      "loss": 4.5873,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799643007593869e-05,
+      "loss": 4.5864,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798804412842817e-05,
+      "loss": 4.5927,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797965818091765e-05,
+      "loss": 4.5817,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797127223340713e-05,
+      "loss": 4.5869,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796288628589661e-05,
+      "loss": 4.5824,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795450033838609e-05,
+      "loss": 4.5824,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794611439087557e-05,
+      "loss": 4.5724,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793772844336505e-05,
+      "loss": 4.5773,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792934249585453e-05,
+      "loss": 4.5723,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792095654834401e-05,
+      "loss": 4.5749,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791257060083349e-05,
+      "loss": 4.5789,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790418465332297e-05,
+      "loss": 4.5645,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789581508461618e-05,
+      "loss": 4.5674,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5644,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787905956839887e-05,
+      "loss": 4.5626,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787067362088835e-05,
+      "loss": 4.5564,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786228767337783e-05,
+      "loss": 4.5581,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785390172586731e-05,
+      "loss": 4.5614,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784553215716052e-05,
+      "loss": 4.5513,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783714620965e-05,
+      "loss": 4.5545,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782876026213948e-05,
+      "loss": 4.5494,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782037431462896e-05,
+      "loss": 4.5409,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7811988367118434e-05,
+      "loss": 4.5633,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5465,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779523285090114e-05,
+      "loss": 4.5533,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786863282194346e-05,
+      "loss": 4.5497,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778477334683826e-05,
+      "loss": 4.5589,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770091387173306e-05,
+      "loss": 4.5495,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761705439662786e-05,
+      "loss": 4.5353,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775331949215226e-05,
+      "loss": 4.5462,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774493354464174e-05,
+      "loss": 4.5517,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773654759713122e-05,
+      "loss": 4.5364,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77281616496207e-05,
+      "loss": 4.5401,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771979208091391e-05,
+      "loss": 4.5343,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771140613340339e-05,
+      "loss": 4.5426,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703020185892875e-05,
+      "loss": 4.5301,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694634238382355e-05,
+      "loss": 4.5373,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686264669675564e-05,
+      "loss": 4.5279,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7677878722165044e-05,
+      "loss": 4.5341,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7669492774654524e-05,
+      "loss": 4.5227,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7661106827144004e-05,
+      "loss": 4.5453,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652720879633484e-05,
+      "loss": 4.5349,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644334932122964e-05,
+      "loss": 4.52,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7635948984612444e-05,
+      "loss": 4.5239,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762757941590565e-05,
+      "loss": 4.5253,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761919346839513e-05,
+      "loss": 4.5294,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761080752088461e-05,
+      "loss": 4.5374,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760242157337409e-05,
+      "loss": 4.5186,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759403562586357e-05,
+      "loss": 4.5237,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758564967835306e-05,
+      "loss": 4.5207,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5229,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.5139,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756050821462523e-05,
+      "loss": 4.5228,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755212226711471e-05,
+      "loss": 4.513,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543769077211646e-05,
+      "loss": 4.5074,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535383129701126e-05,
+      "loss": 4.5209,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7526997182190606e-05,
+      "loss": 4.5128,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518611234680086e-05,
+      "loss": 4.5113,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510225287169566e-05,
+      "loss": 4.5144,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501839339659046e-05,
+      "loss": 4.5057,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.476215839385986,
+      "eval_runtime": 296.0227,
+      "eval_samples_per_second": 1289.06,
+      "eval_steps_per_second": 40.284,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493453392148526e-05,
+      "loss": 4.4964,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748506744463801e-05,
+      "loss": 4.4979,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747668149712749e-05,
+      "loss": 4.51,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746829554961697e-05,
+      "loss": 4.4999,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459909602106446e-05,
+      "loss": 4.51,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451523654595926e-05,
+      "loss": 4.4954,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443137707085406e-05,
+      "loss": 4.4992,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434751759574886e-05,
+      "loss": 4.4932,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426365812064366e-05,
+      "loss": 4.5048,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.4918,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.4915,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.4914,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392838400826015e-05,
+      "loss": 4.4894,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384452453315495e-05,
+      "loss": 4.4841,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376066505804975e-05,
+      "loss": 4.4854,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736768055829446e-05,
+      "loss": 4.4751,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735929461078394e-05,
+      "loss": 4.4805,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735090866327342e-05,
+      "loss": 4.4761,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73425227157629e-05,
+      "loss": 4.4729,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733413676825238e-05,
+      "loss": 4.4989,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732576719954559e-05,
+      "loss": 4.479,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731738125203507e-05,
+      "loss": 4.4748,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730899530452455e-05,
+      "loss": 4.4789,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730060935701403e-05,
+      "loss": 4.4933,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729223978830724e-05,
+      "loss": 4.4671,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728385384079672e-05,
+      "loss": 4.4709,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72754678932862e-05,
+      "loss": 4.4707,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726708194577568e-05,
+      "loss": 4.4775,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725869599826516e-05,
+      "loss": 4.4592,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725031005075464e-05,
+      "loss": 4.4656,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724192410324412e-05,
+      "loss": 4.4678,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72335381557336e-05,
+      "loss": 4.4745,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722515220822308e-05,
+      "loss": 4.4796,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721678263951629e-05,
+      "loss": 4.4692,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720839669200577e-05,
+      "loss": 4.4626,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720001074449525e-05,
+      "loss": 4.4718,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719162479698473e-05,
+      "loss": 4.4509,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718323884947421e-05,
+      "loss": 4.4637,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717485290196369e-05,
+      "loss": 4.4459,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71664833332569e-05,
+      "loss": 4.4526,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158097385746384e-05,
+      "loss": 4.4548,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149711438235864e-05,
+      "loss": 4.4638,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141325490725343e-05,
+      "loss": 4.4538,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7132939543214823e-05,
+      "loss": 4.4517,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124553595704303e-05,
+      "loss": 4.4531,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711616764819378e-05,
+      "loss": 4.4486,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710778170068326e-05,
+      "loss": 4.4614,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70994285107802e-05,
+      "loss": 4.4412,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.431,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708265661575916e-05,
+      "loss": 4.4559,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707427066824864e-05,
+      "loss": 4.4506,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706588472073812e-05,
+      "loss": 4.4405,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70574987732276e-05,
+      "loss": 4.44,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704912920452082e-05,
+      "loss": 4.4428,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70407432570103e-05,
+      "loss": 4.4318,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703235730949978e-05,
+      "loss": 4.4518,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702397136198926e-05,
+      "loss": 4.432,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701558541447874e-05,
+      "loss": 4.4392,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700719946696822e-05,
+      "loss": 4.4324,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69988135194577e-05,
+      "loss": 4.4331,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990443950750906e-05,
+      "loss": 4.4335,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982058003240386e-05,
+      "loss": 4.4316,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973672055729866e-05,
+      "loss": 4.4238,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965302487023075e-05,
+      "loss": 4.4269,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4403,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948530592002035e-05,
+      "loss": 4.4267,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940144644491515e-05,
+      "loss": 4.4232,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931758696980995e-05,
+      "loss": 4.4191,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923372749470475e-05,
+      "loss": 4.4201,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6914986801959955e-05,
+      "loss": 4.4415,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690661723325317e-05,
+      "loss": 4.4354,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898231285742644e-05,
+      "loss": 4.4231,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6889845338232124e-05,
+      "loss": 4.4328,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881459390721604e-05,
+      "loss": 4.4375,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873073443211084e-05,
+      "loss": 4.4212,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864687495700564e-05,
+      "loss": 4.4259,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856301548190044e-05,
+      "loss": 4.4203,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684793197948325e-05,
+      "loss": 4.4105,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683954603197274e-05,
+      "loss": 4.4167,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683116008446222e-05,
+      "loss": 4.4128,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68227741369517e-05,
+      "loss": 4.42,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681438818944118e-05,
+      "loss": 4.4327,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680600224193066e-05,
+      "loss": 4.429,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679761629442014e-05,
+      "loss": 4.4023,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678923034690962e-05,
+      "loss": 4.4198,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678086077820283e-05,
+      "loss": 4.416,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677247483069231e-05,
+      "loss": 4.4188,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676408888318179e-05,
+      "loss": 4.4118,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755719314475e-05,
+      "loss": 4.4044,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674733336696448e-05,
+      "loss": 4.4119,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673894741945396e-05,
+      "loss": 4.4094,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673056147194344e-05,
+      "loss": 4.4049,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672219190323665e-05,
+      "loss": 4.4104,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671380595572613e-05,
+      "loss": 4.4042,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670542000821561e-05,
+      "loss": 4.4095,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669703406070509e-05,
+      "loss": 4.3983,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668864811319457e-05,
+      "loss": 4.4009,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668026216568405e-05,
+      "loss": 4.3992,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667187621817353e-05,
+      "loss": 4.4075,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666350664946674e-05,
+      "loss": 4.4067,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665512070195622e-05,
+      "loss": 4.3931,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66467347544457e-05,
+      "loss": 4.3971,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638348806935175e-05,
+      "loss": 4.3958,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662996285942466e-05,
+      "loss": 4.3972,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662157691191414e-05,
+      "loss": 4.3863,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.3885,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660482139569683e-05,
+      "loss": 4.3991,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659643544818631e-05,
+      "loss": 4.391,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658804950067579e-05,
+      "loss": 4.3873,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657966355316527e-05,
+      "loss": 4.3856,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657127760565475e-05,
+      "loss": 4.3794,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656289165814423e-05,
+      "loss": 4.3972,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655452208943744e-05,
+      "loss": 4.3897,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654615252073065e-05,
+      "loss": 4.3851,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653776657322013e-05,
+      "loss": 4.3887,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529380625709615e-05,
+      "loss": 4.4018,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6520994678199095e-05,
+      "loss": 4.3901,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512608730688575e-05,
+      "loss": 4.3772,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504222783178055e-05,
+      "loss": 4.3875,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495836835667535e-05,
+      "loss": 4.3912,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487450888157015e-05,
+      "loss": 4.3814,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479081319450224e-05,
+      "loss": 4.3844,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470695371939704e-05,
+      "loss": 4.3795,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646232580323291e-05,
+      "loss": 4.3883,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645393985572239e-05,
+      "loss": 4.3748,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644555390821187e-05,
+      "loss": 4.3865,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643716796070135e-05,
+      "loss": 4.3691,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642878201319083e-05,
+      "loss": 4.3849,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642039606568031e-05,
+      "loss": 4.368,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64120101181698e-05,
+      "loss": 4.3928,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640362417065928e-05,
+      "loss": 4.3853,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639523822314876e-05,
+      "loss": 4.3749,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638686865444197e-05,
+      "loss": 4.3695,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637848270693145e-05,
+      "loss": 4.3771,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637009675942093e-05,
+      "loss": 4.3769,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636171081191041e-05,
+      "loss": 4.3875,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635332486439989e-05,
+      "loss": 4.3679,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634493891688936e-05,
+      "loss": 4.3778,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633655296937884e-05,
+      "loss": 4.3738,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632816702186832e-05,
+      "loss": 4.3809,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63197810743578e-05,
+      "loss": 4.3641,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631141150565102e-05,
+      "loss": 4.377,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630304193694423e-05,
+      "loss": 4.3699,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629465598943371e-05,
+      "loss": 4.3633,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628627004192319e-05,
+      "loss": 4.3757,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277884094412666e-05,
+      "loss": 4.3724,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269498146902146e-05,
+      "loss": 4.3624,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261112199391626e-05,
+      "loss": 4.374,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252726251881106e-05,
+      "loss": 4.3637,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.342019557952881,
+      "eval_runtime": 293.1681,
+      "eval_samples_per_second": 1301.612,
+      "eval_steps_per_second": 40.676,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244356683174315e-05,
+      "loss": 4.3607,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6235970735663795e-05,
+      "loss": 4.3545,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227584788153275e-05,
+      "loss": 4.3654,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219198840642755e-05,
+      "loss": 4.3622,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210812893132235e-05,
+      "loss": 4.3764,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202426945621715e-05,
+      "loss": 4.3528,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61940409981112e-05,
+      "loss": 4.3607,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618565505060068e-05,
+      "loss": 4.3528,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617728548189389e-05,
+      "loss": 4.3715,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616889953438337e-05,
+      "loss": 4.3513,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616051358687285e-05,
+      "loss": 4.3529,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615212763936233e-05,
+      "loss": 4.3591,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614375807065554e-05,
+      "loss": 4.3498,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613537212314502e-05,
+      "loss": 4.3495,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61269861756345e-05,
+      "loss": 4.3491,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611860022812398e-05,
+      "loss": 4.3439,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611021428061346e-05,
+      "loss": 4.3416,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610182833310294e-05,
+      "loss": 4.3449,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609344238559242e-05,
+      "loss": 4.3396,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60850564380819e-05,
+      "loss": 4.3674,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076686869375116e-05,
+      "loss": 4.3454,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068300921864595e-05,
+      "loss": 4.345,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3488,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051545405647285e-05,
+      "loss": 4.3596,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043159458136764e-05,
+      "loss": 4.3388,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6034773510626244e-05,
+      "loss": 4.3421,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026403941919454e-05,
+      "loss": 4.3403,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018017994408933e-05,
+      "loss": 4.3454,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009632046898413e-05,
+      "loss": 4.3334,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001246099387893e-05,
+      "loss": 4.3387,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599286015187737e-05,
+      "loss": 4.3426,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598447420436685e-05,
+      "loss": 4.3464,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597610463566007e-05,
+      "loss": 4.3523,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596771868814955e-05,
+      "loss": 4.3442,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595933274063903e-05,
+      "loss": 4.3373,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59509467931285e-05,
+      "loss": 4.3432,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594256084561798e-05,
+      "loss": 4.3287,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593417489810746e-05,
+      "loss": 4.3363,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592578895059694e-05,
+      "loss": 4.3226,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591740300308642e-05,
+      "loss": 4.327,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590903343437963e-05,
+      "loss": 4.3311,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590064748686911e-05,
+      "loss": 4.3443,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589226153935859e-05,
+      "loss": 4.3356,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588387559184808e-05,
+      "loss": 4.3273,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587550602314129e-05,
+      "loss": 4.3315,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867136454434496e-05,
+      "loss": 4.3272,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858750506923976e-05,
+      "loss": 4.339,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850364559413456e-05,
+      "loss": 4.3211,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5841978611902936e-05,
+      "loss": 4.3117,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833592664392416e-05,
+      "loss": 4.335,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825206716881896e-05,
+      "loss": 4.3308,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816837148175105e-05,
+      "loss": 4.317,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808451200664585e-05,
+      "loss": 4.3255,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800065253154065e-05,
+      "loss": 4.3269,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791679305643545e-05,
+      "loss": 4.3115,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783293358133025e-05,
+      "loss": 4.3325,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577490741062251e-05,
+      "loss": 4.3115,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576653784191572e-05,
+      "loss": 4.3266,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57581518944052e-05,
+      "loss": 4.3122,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574976594689468e-05,
+      "loss": 4.3237,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574137999938416e-05,
+      "loss": 4.312,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573299405187364e-05,
+      "loss": 4.3162,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572460810436312e-05,
+      "loss": 4.3046,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57162221568526e-05,
+      "loss": 4.3144,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570783620934208e-05,
+      "loss": 4.3297,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569946664063529e-05,
+      "loss": 4.3113,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569108069312477e-05,
+      "loss": 4.3122,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568269474561425e-05,
+      "loss": 4.3029,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.3089,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665955608200674e-05,
+      "loss": 4.3284,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657569660690154e-05,
+      "loss": 4.3221,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649183713179634e-05,
+      "loss": 4.3152,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640797765669114e-05,
+      "loss": 4.3222,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563242819696232e-05,
+      "loss": 4.3225,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56240422494518e-05,
+      "loss": 4.3118,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561565630194128e-05,
+      "loss": 4.3145,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560727035443076e-05,
+      "loss": 4.3123,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559888440692024e-05,
+      "loss": 4.3034,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559049845940972e-05,
+      "loss": 4.3051,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55821125118992e-05,
+      "loss": 4.3031,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557372656438868e-05,
+      "loss": 4.31,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55653569956819e-05,
+      "loss": 4.3226,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555697104817138e-05,
+      "loss": 4.3194,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554858510066086e-05,
+      "loss": 4.2941,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554019915315033e-05,
+      "loss": 4.3174,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553181320563981e-05,
+      "loss": 4.3088,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552342725812929e-05,
+      "loss": 4.3092,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551504131061877e-05,
+      "loss": 4.3038,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550667174191198e-05,
+      "loss": 4.2965,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549828579440146e-05,
+      "loss": 4.3069,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548989984689094e-05,
+      "loss": 4.3016,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548151389938042e-05,
+      "loss": 4.3005,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54731279518699e-05,
+      "loss": 4.3034,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546474200435939e-05,
+      "loss": 4.2988,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456372435652596e-05,
+      "loss": 4.3089,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5447986488142076e-05,
+      "loss": 4.2961,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439600540631556e-05,
+      "loss": 4.2949,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431214593121036e-05,
+      "loss": 4.2935,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422828645610516e-05,
+      "loss": 4.3071,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414442698099996e-05,
+      "loss": 4.3,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406056750589476e-05,
+      "loss": 4.2911,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397670803078956e-05,
+      "loss": 4.2907,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389301234372165e-05,
+      "loss": 4.2963,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380915286861645e-05,
+      "loss": 4.2986,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372529339351125e-05,
+      "loss": 4.2881,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536415977064434e-05,
+      "loss": 4.2844,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535577382313382e-05,
+      "loss": 4.2982,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53473878756233e-05,
+      "loss": 4.2866,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533900192811278e-05,
+      "loss": 4.2877,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533061598060226e-05,
+      "loss": 4.2837,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532223003309174e-05,
+      "loss": 4.2853,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531384408558122e-05,
+      "loss": 4.2975,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305458138070694e-05,
+      "loss": 4.2845,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297072190560174e-05,
+      "loss": 4.2862,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.2921,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.2997,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527194710563608e-05,
+      "loss": 4.2982,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526356115812556e-05,
+      "loss": 4.2771,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525517521061504e-05,
+      "loss": 4.2894,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524678926310452e-05,
+      "loss": 4.2892,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238403315594e-05,
+      "loss": 4.2877,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230033746887215e-05,
+      "loss": 4.2908,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221647799376694e-05,
+      "loss": 4.2796,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521326185186617e-05,
+      "loss": 4.2908,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520487590435565e-05,
+      "loss": 4.2777,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519648995684513e-05,
+      "loss": 4.2888,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518810400933461e-05,
+      "loss": 4.2738,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179734440627817e-05,
+      "loss": 4.2902,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171348493117297e-05,
+      "loss": 4.2718,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5162962545606777e-05,
+      "loss": 4.3017,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154576598096256e-05,
+      "loss": 4.2832,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514619065058574e-05,
+      "loss": 4.2839,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513780470307522e-05,
+      "loss": 4.2733,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51294187555647e-05,
+      "loss": 4.2856,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512103280805418e-05,
+      "loss": 4.2824,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511266323934739e-05,
+      "loss": 4.293,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510427729183687e-05,
+      "loss": 4.2757,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509589134432635e-05,
+      "loss": 4.284,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508752177561956e-05,
+      "loss": 4.2806,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507913582810904e-05,
+      "loss": 4.2924,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507074988059852e-05,
+      "loss": 4.2686,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062363933088e-05,
+      "loss": 4.2843,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505399436438121e-05,
+      "loss": 4.2766,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50456084168707e-05,
+      "loss": 4.2751,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503722246936018e-05,
+      "loss": 4.2836,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502883652184966e-05,
+      "loss": 4.2806,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502045057433914e-05,
+      "loss": 4.2721,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501206462682862e-05,
+      "loss": 4.2812,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50036786793181e-05,
+      "loss": 4.2767,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.258913040161133,
+      "eval_runtime": 293.9356,
+      "eval_samples_per_second": 1298.213,
+      "eval_steps_per_second": 40.57,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995309110611306e-05,
+      "loss": 4.271,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986923163100786e-05,
+      "loss": 4.261,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978537215590266e-05,
+      "loss": 4.276,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970151268079746e-05,
+      "loss": 4.2723,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961765320569226e-05,
+      "loss": 4.2868,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953379373058706e-05,
+      "loss": 4.2678,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494499342554818e-05,
+      "loss": 4.2673,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936607478037665e-05,
+      "loss": 4.2653,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928221530527145e-05,
+      "loss": 4.2806,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919835583016625e-05,
+      "loss": 4.2678,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911449635506105e-05,
+      "loss": 4.264,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903063687995585e-05,
+      "loss": 4.2678,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.2617,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886308171778274e-05,
+      "loss": 4.2644,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877922224267754e-05,
+      "loss": 4.2674,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869536276757234e-05,
+      "loss": 4.2494,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861150329246714e-05,
+      "loss": 4.2585,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852764381736194e-05,
+      "loss": 4.2578,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844378434225674e-05,
+      "loss": 4.2528,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4835992486715154e-05,
+      "loss": 4.2768,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827606539204634e-05,
+      "loss": 4.2656,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2578,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2633,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480246507547681e-05,
+      "loss": 4.2751,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479407912796629e-05,
+      "loss": 4.254,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478569318045577e-05,
+      "loss": 4.2588,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477730723294524e-05,
+      "loss": 4.2559,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476892128543472e-05,
+      "loss": 4.258,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47605353379242e-05,
+      "loss": 4.2497,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2589,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2591,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2629,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.2691,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471863835797907e-05,
+      "loss": 4.2605,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471025241046855e-05,
+      "loss": 4.2527,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470186646295803e-05,
+      "loss": 4.2598,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469348051544751e-05,
+      "loss": 4.2512,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685110946740717e-05,
+      "loss": 4.2521,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676724999230197e-05,
+      "loss": 4.2426,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668339051719676e-05,
+      "loss": 4.2464,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4659953104209156e-05,
+      "loss": 4.2502,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651583535502366e-05,
+      "loss": 4.2612,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2516,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634811640481325e-05,
+      "loss": 4.2485,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626425692970805e-05,
+      "loss": 4.2508,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.248,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46096701767535e-05,
+      "loss": 4.2582,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460128422924298e-05,
+      "loss": 4.242,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.2303,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458451233422194e-05,
+      "loss": 4.2539,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457614276551515e-05,
+      "loss": 4.2521,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456775681800463e-05,
+      "loss": 4.237,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455937087049411e-05,
+      "loss": 4.2447,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455098492298359e-05,
+      "loss": 4.249,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454259897547307e-05,
+      "loss": 4.2321,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2534,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452584345925576e-05,
+      "loss": 4.2351,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451745751174524e-05,
+      "loss": 4.2466,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450907156423472e-05,
+      "loss": 4.2353,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500685616724206e-05,
+      "loss": 4.2441,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.2367,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483930100506895e-05,
+      "loss": 4.242,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475544152996375e-05,
+      "loss": 4.2262,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.2404,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458772257975335e-05,
+      "loss": 4.2489,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450386310464815e-05,
+      "loss": 4.2369,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442000362954295e-05,
+      "loss": 4.2325,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433614415443775e-05,
+      "loss": 4.2262,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425228467933254e-05,
+      "loss": 4.2318,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441684252042273e-05,
+      "loss": 4.2483,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440845657291221e-05,
+      "loss": 4.2471,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440007062540169e-05,
+      "loss": 4.2405,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391684677891174e-05,
+      "loss": 4.2463,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438331510918438e-05,
+      "loss": 4.2485,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2336,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.2356,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435819002426029e-05,
+      "loss": 4.2407,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434980407674977e-05,
+      "loss": 4.2285,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434141812923925e-05,
+      "loss": 4.2274,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433303218172873e-05,
+      "loss": 4.231,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43246462342182e-05,
+      "loss": 4.2327,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431626028670768e-05,
+      "loss": 4.2489,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430787433919716e-05,
+      "loss": 4.2474,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429948839168664e-05,
+      "loss": 4.2216,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429110244417613e-05,
+      "loss": 4.2394,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428271649666561e-05,
+      "loss": 4.2358,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427433054915509e-05,
+      "loss": 4.2357,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426594460164457e-05,
+      "loss": 4.2323,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425757503293778e-05,
+      "loss": 4.2224,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424918908542726e-05,
+      "loss": 4.2309,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424080313791674e-05,
+      "loss": 4.2294,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423241719040622e-05,
+      "loss": 4.2283,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42240312428957e-05,
+      "loss": 4.2266,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215661674188906e-05,
+      "loss": 4.231,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207275726678386e-05,
+      "loss": 4.233,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198889779167866e-05,
+      "loss": 4.2309,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190503831657346e-05,
+      "loss": 4.2195,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182117884146826e-05,
+      "loss": 4.2255,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.231,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416536236792952e-05,
+      "loss": 4.2326,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4156976420419e-05,
+      "loss": 4.2192,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414859047290848e-05,
+      "loss": 4.2226,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414020452539796e-05,
+      "loss": 4.2267,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413183495669117e-05,
+      "loss": 4.2262,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412346538798438e-05,
+      "loss": 4.2196,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411507944047386e-05,
+      "loss": 4.2157,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410669349296334e-05,
+      "loss": 4.2235,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409830754545282e-05,
+      "loss": 4.2176,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.2161,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408153565043178e-05,
+      "loss": 4.2146,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407314970292126e-05,
+      "loss": 4.2151,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406476375541074e-05,
+      "loss": 4.2273,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2179,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.2144,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039638670486644e-05,
+      "loss": 4.2235,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.2325,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022866775465604e-05,
+      "loss": 4.229,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014480827955084e-05,
+      "loss": 4.2089,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400609488044456e-05,
+      "loss": 4.2173,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399770893293404e-05,
+      "loss": 4.2214,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398932298542352e-05,
+      "loss": 4.2164,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398095341671673e-05,
+      "loss": 4.2243,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397256746920621e-05,
+      "loss": 4.211,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396418152169569e-05,
+      "loss": 4.2196,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395579557418517e-05,
+      "loss": 4.2168,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394740962667465e-05,
+      "loss": 4.2198,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393902367916413e-05,
+      "loss": 4.2063,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393065411045734e-05,
+      "loss": 4.2193,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392226816294682e-05,
+      "loss": 4.2123,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39138822154363e-05,
+      "loss": 4.2262,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390549626792578e-05,
+      "loss": 4.2143,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389711032041526e-05,
+      "loss": 4.221,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388872437290474e-05,
+      "loss": 4.206,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388035480419795e-05,
+      "loss": 4.2219,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387196885668744e-05,
+      "loss": 4.2172,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386358290917692e-05,
+      "loss": 4.2235,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38551969616664e-05,
+      "loss": 4.2068,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384681101415588e-05,
+      "loss": 4.2227,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383842506664536e-05,
+      "loss": 4.211,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383003911913484e-05,
+      "loss": 4.2268,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821669550428046e-05,
+      "loss": 4.2077,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813283602917526e-05,
+      "loss": 4.2185,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804897655407006e-05,
+      "loss": 4.2112,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796511707896486e-05,
+      "loss": 4.2068,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788125760385966e-05,
+      "loss": 4.2182,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779739812875446e-05,
+      "loss": 4.2147,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377135386536492e-05,
+      "loss": 4.2068,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3762984296658135e-05,
+      "loss": 4.2136,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375459834914762e-05,
+      "loss": 4.215,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2025299072265625,
+      "eval_runtime": 295.0207,
+      "eval_samples_per_second": 1293.438,
+      "eval_steps_per_second": 40.421,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37462124016371e-05,
+      "loss": 4.2057,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3737826454126575e-05,
+      "loss": 4.199,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729440506616055e-05,
+      "loss": 4.2105,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721054559105535e-05,
+      "loss": 4.2082,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3712668611595015e-05,
+      "loss": 4.2192,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704282664084495e-05,
+      "loss": 4.2062,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3695896716573975e-05,
+      "loss": 4.203,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687510769063455e-05,
+      "loss": 4.2031,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3679124821552935e-05,
+      "loss": 4.2163,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3670738874042415e-05,
+      "loss": 4.2032,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3662352926531895e-05,
+      "loss": 4.2038,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3653966979021375e-05,
+      "loss": 4.2022,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3645597410314584e-05,
+      "loss": 4.2015,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363721146280407e-05,
+      "loss": 4.2026,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362882551529355e-05,
+      "loss": 4.2039,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362043956778303e-05,
+      "loss": 4.1886,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361205362027251e-05,
+      "loss": 4.1932,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360366767276199e-05,
+      "loss": 4.198,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595281725251463e-05,
+      "loss": 4.1887,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358689577774094e-05,
+      "loss": 4.2128,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357850983023042e-05,
+      "loss": 4.2054,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35701238827199e-05,
+      "loss": 4.1999,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356173793520938e-05,
+      "loss": 4.1956,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355335198769886e-05,
+      "loss": 4.2112,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354498241899207e-05,
+      "loss": 4.1965,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353659647148155e-05,
+      "loss": 4.1988,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352821052397104e-05,
+      "loss": 4.1903,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351984095526425e-05,
+      "loss": 4.1968,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351145500775373e-05,
+      "loss": 4.1937,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350306906024321e-05,
+      "loss": 4.1933,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349468311273269e-05,
+      "loss": 4.1983,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348629716522217e-05,
+      "loss": 4.202,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347792759651538e-05,
+      "loss": 4.205,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346954164900486e-05,
+      "loss": 4.2025,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346115570149434e-05,
+      "loss": 4.1906,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345276975398382e-05,
+      "loss": 4.2012,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444400185277026e-05,
+      "loss": 4.1934,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436014237766506e-05,
+      "loss": 4.1836,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342762829025599e-05,
+      "loss": 4.1882,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341924234274547e-05,
+      "loss": 4.1857,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341085639523495e-05,
+      "loss": 4.1894,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340247044772443e-05,
+      "loss": 4.2037,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339408450021391e-05,
+      "loss": 4.1889,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338569855270339e-05,
+      "loss": 4.1902,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33773289839966e-05,
+      "loss": 4.1968,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336894303648608e-05,
+      "loss": 4.1877,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336055708897556e-05,
+      "loss": 4.1941,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335217114146504e-05,
+      "loss": 4.1835,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334378519395452e-05,
+      "loss": 4.1707,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333541562524773e-05,
+      "loss": 4.1988,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332702967773721e-05,
+      "loss": 4.1894,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331864373022669e-05,
+      "loss": 4.1842,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331025778271618e-05,
+      "loss": 4.1843,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3301888214009386e-05,
+      "loss": 4.1865,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293502266498866e-05,
+      "loss": 4.1725,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285132697792075e-05,
+      "loss": 4.1911,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3276746750281555e-05,
+      "loss": 4.1819,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268360802771035e-05,
+      "loss": 4.1873,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3259974855260515e-05,
+      "loss": 4.1806,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251588907749995e-05,
+      "loss": 4.1836,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324320296023947e-05,
+      "loss": 4.178,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323481701272895e-05,
+      "loss": 4.1869,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322643106521843e-05,
+      "loss": 4.1723,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218061496511644e-05,
+      "loss": 4.1811,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3209675549001124e-05,
+      "loss": 4.1904,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3201289601490604e-05,
+      "loss": 4.1814,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3192903653980084e-05,
+      "loss": 4.1754,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31845340852733e-05,
+      "loss": 4.1732,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317614813776277e-05,
+      "loss": 4.1728,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316776219025225e-05,
+      "loss": 4.1883,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315937624274173e-05,
+      "loss": 4.192,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315100667403494e-05,
+      "loss": 4.1832,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314262072652442e-05,
+      "loss": 4.1898,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31342347790139e-05,
+      "loss": 4.1912,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312584883150338e-05,
+      "loss": 4.1848,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31174792627966e-05,
+      "loss": 4.1767,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109109694089814e-05,
+      "loss": 4.18,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3100723746579294e-05,
+      "loss": 4.1777,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092337799068773e-05,
+      "loss": 4.1688,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308395185155825e-05,
+      "loss": 4.1797,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307556590404773e-05,
+      "loss": 4.1733,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067179956537207e-05,
+      "loss": 4.1901,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058794009026687e-05,
+      "loss": 4.1935,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050408061516166e-05,
+      "loss": 4.164,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042022114005646e-05,
+      "loss": 4.1863,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3033636166495126e-05,
+      "loss": 4.1761,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025250218984606e-05,
+      "loss": 4.1803,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016880650277815e-05,
+      "loss": 4.1762,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30084947027673e-05,
+      "loss": 4.1686,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300010875525678e-05,
+      "loss": 4.177,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299172280774626e-05,
+      "loss": 4.1733,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298333686023574e-05,
+      "loss": 4.1743,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297495091272522e-05,
+      "loss": 4.1707,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29665649652147e-05,
+      "loss": 4.1784,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295817901770418e-05,
+      "loss": 4.1778,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294980944899739e-05,
+      "loss": 4.1754,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294142350148687e-05,
+      "loss": 4.1652,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293303755397635e-05,
+      "loss": 4.1721,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292465160646583e-05,
+      "loss": 4.1739,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2916265658955304e-05,
+      "loss": 4.1802,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290789609024852e-05,
+      "loss": 4.1655,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2899510142738e-05,
+      "loss": 4.1709,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2891124195227487e-05,
+      "loss": 4.1672,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288273824771696e-05,
+      "loss": 4.175,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287435230020644e-05,
+      "loss": 4.1623,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286596635269592e-05,
+      "loss": 4.1667,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285759678398913e-05,
+      "loss": 4.1679,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284921083647861e-05,
+      "loss": 4.1665,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284082488896809e-05,
+      "loss": 4.1616,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283243894145757e-05,
+      "loss": 4.1667,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282406937275078e-05,
+      "loss": 4.1618,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281568342524026e-05,
+      "loss": 4.1686,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280729747772974e-05,
+      "loss": 4.1702,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2798911530219224e-05,
+      "loss": 4.1582,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2790541961512434e-05,
+      "loss": 4.1711,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278217239280565e-05,
+      "loss": 4.1806,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277378644529513e-05,
+      "loss": 4.1781,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276541687658834e-05,
+      "loss": 4.1569,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275703092907782e-05,
+      "loss": 4.165,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27486449815673e-05,
+      "loss": 4.1691,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274025903405678e-05,
+      "loss": 4.1625,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273188946534999e-05,
+      "loss": 4.1723,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272350351783947e-05,
+      "loss": 4.1584,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271511757032895e-05,
+      "loss": 4.1696,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270673162281843e-05,
+      "loss": 4.1629,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269834567530791e-05,
+      "loss": 4.1714,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268995972779739e-05,
+      "loss": 4.157,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268157378028687e-05,
+      "loss": 4.1625,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267318783277635e-05,
+      "loss": 4.166,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266480188526583e-05,
+      "loss": 4.1733,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656432316559036e-05,
+      "loss": 4.1664,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648046369048516e-05,
+      "loss": 4.1661,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639660421537996e-05,
+      "loss": 4.1561,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2631274474027476e-05,
+      "loss": 4.1715,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622888526516956e-05,
+      "loss": 4.1653,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614502579006436e-05,
+      "loss": 4.1737,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606133010299645e-05,
+      "loss": 4.1601,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259774706278913e-05,
+      "loss": 4.1706,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258936111527861e-05,
+      "loss": 4.1599,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258097516776809e-05,
+      "loss": 4.1809,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257258922025757e-05,
+      "loss": 4.155,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256420327274705e-05,
+      "loss": 4.1712,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255581732523653e-05,
+      "loss": 4.1609,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254743137772601e-05,
+      "loss": 4.1597,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253906180901922e-05,
+      "loss": 4.1672,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25306758615087e-05,
+      "loss": 4.1612,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252228991399818e-05,
+      "loss": 4.1651,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251390396648766e-05,
+      "loss": 4.1585,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250553439778087e-05,
+      "loss": 4.1682,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.161040306091309,
+      "eval_runtime": 311.0389,
+      "eval_samples_per_second": 1226.827,
+      "eval_steps_per_second": 38.339,
+      "step": 457920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.8787776957255702e+17,
+  "trial_name": null,
+  "trial_params": null
+}