diff --git "a/checkpoint-1221120/trainer_state.json" "b/checkpoint-1221120/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1221120/trainer_state.json"
@@ -0,0 +1,14463 @@
+{
+  "best_metric": 4.015658855438232,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-only/lstm/4/checkpoints/checkpoint-1221120",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1221120,
+  "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": 10.8184,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5451,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0556,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9887,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9431,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8973,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7348,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6336,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5273,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4566,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3977,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3308,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2558,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1955,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.145,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0798,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0465,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 5.9979,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9481,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.917,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.8822,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8444,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8204,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7788,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7623,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7354,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7112,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6819,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.649,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6384,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6163,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6035,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.577,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5583,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5497,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5321,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.514,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4821,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4781,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4476,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4493,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4229,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4256,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3959,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3891,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3828,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3748,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3582,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3361,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3333,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3368,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3252,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3036,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2917,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2894,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.276,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2855,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2423,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2558,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2417,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2278,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2337,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.212,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.2028,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1976,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.2094,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1848,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9438288926028735e-05,
+      "loss": 5.1791,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1638,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1566,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1608,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404761514790384e-05,
+      "loss": 5.1653,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1378,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1479,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1487,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.128,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1115,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1112,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.1057,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.0933,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932932074480317e-05,
+      "loss": 5.0926,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0956,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0931,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.076,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0558,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0611,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0656,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0539,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0513,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0326,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0308,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0253,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0247,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0243,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0073,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0246,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0087,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9881,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9849,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.9853,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9819,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9812,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9737,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9671,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9612,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.942,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9451,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9444,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9388,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9337,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9407,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.9148,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.9172,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052682749778396e-05,
+      "loss": 4.9131,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9185,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9223,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9046,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9021,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.8934,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.8998,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.8951,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8896,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8815,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8769,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8693,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8597,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8653,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.859,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892694267353179e-05,
+      "loss": 4.8529,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8453,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8602,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8606,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8505,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8375,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8365,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8403,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.836,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8328,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.8295,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8242,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8182,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.814,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8165,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8043,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8004,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8084,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.7941,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.8023,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7922,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7842,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.752739429473877,
+      "eval_runtime": 300.3691,
+      "eval_samples_per_second": 1270.407,
+      "eval_steps_per_second": 39.701,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.778,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7784,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7878,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7773,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7801,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7651,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7681,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7531,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7649,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7623,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7686,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.7492,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7493,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7438,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7341,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7381,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.7463,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7295,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7312,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7444,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7316,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7323,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7185,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.733,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.7182,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.7158,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.7115,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.7073,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.6985,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7007,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.711,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7044,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.698,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.7066,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.6985,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.6957,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.6826,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.6847,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.6739,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.6851,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840721047352432e-05,
+      "loss": 4.6692,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83988245260138e-05,
+      "loss": 4.6829,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.6646,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838205263099276e-05,
+      "loss": 4.6716,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837368306228597e-05,
+      "loss": 4.6707,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836529711477545e-05,
+      "loss": 4.6694,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.665,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6523,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6504,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.6657,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6582,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.653,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6409,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6514,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6387,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6634,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.6263,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.6481,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.6374,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6274,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6397,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6336,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.618,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6249,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6379,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.622,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189257732269455e-05,
+      "loss": 4.6185,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180871784758935e-05,
+      "loss": 4.6172,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172485837248415e-05,
+      "loss": 4.6109,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164099889737895e-05,
+      "loss": 4.6203,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155713942227375e-05,
+      "loss": 4.6323,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6078,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.6267,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6263,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6145,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.5999,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6122,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.608,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.5998,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6003,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807188722472964e-05,
+      "loss": 4.6037,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806351765602285e-05,
+      "loss": 4.6125,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805513170851233e-05,
+      "loss": 4.6023,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804674576100181e-05,
+      "loss": 4.5851,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.5949,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.5958,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.5931,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013218349763464e-05,
+      "loss": 4.5984,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004832402252944e-05,
+      "loss": 4.5839,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799646283354615e-05,
+      "loss": 4.5811,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798807688603563e-05,
+      "loss": 4.5808,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797969093852511e-05,
+      "loss": 4.5854,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797130499101459e-05,
+      "loss": 4.5825,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796291904350407e-05,
+      "loss": 4.5769,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954533095993546e-05,
+      "loss": 4.5957,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946147148483026e-05,
+      "loss": 4.5833,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793777757977624e-05,
+      "loss": 4.5664,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792939163226572e-05,
+      "loss": 4.5694,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5723,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5721,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5704,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.5714,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5652,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.5674,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5462,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.5595,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78539508622785e-05,
+      "loss": 4.5533,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784556491476798e-05,
+      "loss": 4.5564,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783717896725746e-05,
+      "loss": 4.5502,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.5608,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782040707223642e-05,
+      "loss": 4.5423,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812021124725906e-05,
+      "loss": 4.5444,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803635177215386e-05,
+      "loss": 4.5474,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795249229704866e-05,
+      "loss": 4.5468,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786863282194346e-05,
+      "loss": 4.5611,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778477334683826e-05,
+      "loss": 4.5491,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770091387173306e-05,
+      "loss": 4.5464,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761705439662786e-05,
+      "loss": 4.5389,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775331949215226e-05,
+      "loss": 4.5455,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744966302249204e-05,
+      "loss": 4.5448,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736580354738684e-05,
+      "loss": 4.5375,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728194407228164e-05,
+      "loss": 4.5388,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719808459717644e-05,
+      "loss": 4.5349,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711422512207124e-05,
+      "loss": 4.5331,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703036564696604e-05,
+      "loss": 4.5238,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694650617186084e-05,
+      "loss": 4.5302,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686264669675564e-05,
+      "loss": 4.5254,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7677878722165044e-05,
+      "loss": 4.5294,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766950915345826e-05,
+      "loss": 4.5158,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766112320594773e-05,
+      "loss": 4.538,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765273725843721e-05,
+      "loss": 4.5424,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764435131092669e-05,
+      "loss": 4.5273,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.521,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762761217351312e-05,
+      "loss": 4.5208,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76192262260026e-05,
+      "loss": 4.5256,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761084027849208e-05,
+      "loss": 4.5264,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760245433098156e-05,
+      "loss": 4.5213,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759406838347104e-05,
+      "loss": 4.5206,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758568243596052e-05,
+      "loss": 4.5163,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757729648845e-05,
+      "loss": 4.521,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756891054093948e-05,
+      "loss": 4.5116,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560540972232686e-05,
+      "loss": 4.5206,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552155024722166e-05,
+      "loss": 4.5118,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543769077211646e-05,
+      "loss": 4.5046,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535383129701126e-05,
+      "loss": 4.5169,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.5045,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5126,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510241665973295e-05,
+      "loss": 4.5018,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750185571846278e-05,
+      "loss": 4.4998,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.470576763153076,
+      "eval_runtime": 296.4242,
+      "eval_samples_per_second": 1287.314,
+      "eval_steps_per_second": 40.23,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749346977095226e-05,
+      "loss": 4.4978,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.4885,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.5106,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.4968,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5068,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.4903,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.4986,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743476813837862e-05,
+      "loss": 4.4792,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426382190868095e-05,
+      "loss": 4.4961,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.4986,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.5,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.4903,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4805,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.4823,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.478,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736769693709819e-05,
+      "loss": 4.479,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735931098958767e-05,
+      "loss": 4.4885,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735092504207715e-05,
+      "loss": 4.4742,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734253909456663e-05,
+      "loss": 4.4759,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733415314705611e-05,
+      "loss": 4.4963,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732576719954559e-05,
+      "loss": 4.4807,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.4836,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730901168332828e-05,
+      "loss": 4.4714,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730062573581776e-05,
+      "loss": 4.4888,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4669,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4738,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4689,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267098324579415e-05,
+      "loss": 4.4696,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258712377068895e-05,
+      "loss": 4.4551,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4626,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4738,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4715,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4673,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216799018320024e-05,
+      "loss": 4.4681,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4658,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4657,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719165755459219e-05,
+      "loss": 4.4534,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718327160708167e-05,
+      "loss": 4.4565,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717488565957115e-05,
+      "loss": 4.4461,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716649971206063e-05,
+      "loss": 4.4633,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715811376455011e-05,
+      "loss": 4.4467,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714972781703959e-05,
+      "loss": 4.4566,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714134186952907e-05,
+      "loss": 4.4478,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713295592201855e-05,
+      "loss": 4.4472,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712456997450803e-05,
+      "loss": 4.4576,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711618402699751e-05,
+      "loss": 4.4491,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710781445829072e-05,
+      "loss": 4.4487,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70994285107802e-05,
+      "loss": 4.4309,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.4382,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4508,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707428704705237e-05,
+      "loss": 4.4422,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.444,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705751515203134e-05,
+      "loss": 4.4285,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704912920452082e-05,
+      "loss": 4.4429,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70407432570103e-05,
+      "loss": 4.4278,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4534,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4203,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4411,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4364,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4219,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4314,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982074382044115e-05,
+      "loss": 4.436,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973688434533595e-05,
+      "loss": 4.4173,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965302487023075e-05,
+      "loss": 4.4255,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4353,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694854697080577e-05,
+      "loss": 4.4285,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694016102329525e-05,
+      "loss": 4.4182,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693177507578473e-05,
+      "loss": 4.423,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692338912827421e-05,
+      "loss": 4.4133,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691500318076369e-05,
+      "loss": 4.4268,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690661723325317e-05,
+      "loss": 4.4399,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898231285742644e-05,
+      "loss": 4.4164,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688986171703586e-05,
+      "loss": 4.4347,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688147576952534e-05,
+      "loss": 4.4323,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687308982201481e-05,
+      "loss": 4.4251,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686472025330803e-05,
+      "loss": 4.4103,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685633430579751e-05,
+      "loss": 4.4273,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684794835828699e-05,
+      "loss": 4.4184,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683956241077647e-05,
+      "loss": 4.4188,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683117646326595e-05,
+      "loss": 4.4143,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682279051575543e-05,
+      "loss": 4.4178,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681440456824491e-05,
+      "loss": 4.432,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680601862073439e-05,
+      "loss": 4.4166,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67976490520276e-05,
+      "loss": 4.4025,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678926310451708e-05,
+      "loss": 4.4101,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678087715700656e-05,
+      "loss": 4.4157,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677249120949604e-05,
+      "loss": 4.4159,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676410526198552e-05,
+      "loss": 4.4198,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755719314475e-05,
+      "loss": 4.4037,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.4056,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673898017706142e-05,
+      "loss": 4.3989,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4071,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.4077,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.4064,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.4234,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.4043,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.4001,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668029492329151e-05,
+      "loss": 4.3971,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667190897578099e-05,
+      "loss": 4.4005,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666352302827047e-05,
+      "loss": 4.4054,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665513708075995e-05,
+      "loss": 4.4013,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664675113324943e-05,
+      "loss": 4.3996,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663836518573891e-05,
+      "loss": 4.3962,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662997923822839e-05,
+      "loss": 4.4017,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662160966952161e-05,
+      "loss": 4.3788,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661322372201109e-05,
+      "loss": 4.394,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604837774500567e-05,
+      "loss": 4.3865,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596468205793776e-05,
+      "loss": 4.3938,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588082258283256e-05,
+      "loss": 4.3883,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579696310772736e-05,
+      "loss": 4.396,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571310363262216e-05,
+      "loss": 4.3799,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562924415751695e-05,
+      "loss": 4.3794,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554538468241175e-05,
+      "loss": 4.3906,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654615252073065e-05,
+      "loss": 4.3827,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653776657322013e-05,
+      "loss": 4.4023,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529397004513344e-05,
+      "loss": 4.3935,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652102743580656e-05,
+      "loss": 4.3909,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651264148829604e-05,
+      "loss": 4.3793,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650425554078552e-05,
+      "loss": 4.3847,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495869593275e-05,
+      "loss": 4.3892,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648748364576447e-05,
+      "loss": 4.385,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.3841,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647072812954717e-05,
+      "loss": 4.3744,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646234218203665e-05,
+      "loss": 4.381,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645395623452612e-05,
+      "loss": 4.3689,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3775,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3718,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.3762,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642042882328778e-05,
+      "loss": 4.3643,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641204287577726e-05,
+      "loss": 4.3886,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640365692826674e-05,
+      "loss": 4.3892,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639527098075622e-05,
+      "loss": 4.3803,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63868850332457e-05,
+      "loss": 4.366,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637849908573518e-05,
+      "loss": 4.3752,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637011313822466e-05,
+      "loss": 4.3741,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636174356951787e-05,
+      "loss": 4.3805,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635335762200735e-05,
+      "loss": 4.373,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634497167449683e-05,
+      "loss": 4.3716,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633658572698631e-05,
+      "loss": 4.3691,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632819977947579e-05,
+      "loss": 4.3776,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319830210769e-05,
+      "loss": 4.3672,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631144426325848e-05,
+      "loss": 4.3736,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630305831574796e-05,
+      "loss": 4.3666,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629467236823744e-05,
+      "loss": 4.3642,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628628642072692e-05,
+      "loss": 4.3738,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62779004732164e-05,
+      "loss": 4.3678,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626951452570588e-05,
+      "loss": 4.3634,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626112857819536e-05,
+      "loss": 4.3607,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625275900948857e-05,
+      "loss": 4.3572,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.336536884307861,
+      "eval_runtime": 295.5726,
+      "eval_samples_per_second": 1291.023,
+      "eval_steps_per_second": 40.345,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624437306197805e-05,
+      "loss": 4.3615,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623598711446753e-05,
+      "loss": 4.3508,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622760116695701e-05,
+      "loss": 4.3687,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219215219446484e-05,
+      "loss": 4.3569,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621082927193597e-05,
+      "loss": 4.3682,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620244332442545e-05,
+      "loss": 4.3531,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619405737691493e-05,
+      "loss": 4.3577,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618568780820814e-05,
+      "loss": 4.3426,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617730186069762e-05,
+      "loss": 4.3591,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61689159131871e-05,
+      "loss": 4.3614,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616052996567658e-05,
+      "loss": 4.3627,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615216039696979e-05,
+      "loss": 4.3555,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614377444945927e-05,
+      "loss": 4.3474,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613538850194875e-05,
+      "loss": 4.3485,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612700255443823e-05,
+      "loss": 4.3421,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611863298573144e-05,
+      "loss": 4.3461,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110247038220925e-05,
+      "loss": 4.3539,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6101861090710405e-05,
+      "loss": 4.3415,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093475143199885e-05,
+      "loss": 4.3444,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085089195689365e-05,
+      "loss": 4.3661,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076703248178845e-05,
+      "loss": 4.3491,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068317300668325e-05,
+      "loss": 4.3496,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3429,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051545405647285e-05,
+      "loss": 4.3557,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043175836940494e-05,
+      "loss": 4.3387,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6034789889429974e-05,
+      "loss": 4.3509,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026403941919454e-05,
+      "loss": 4.3376,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018017994408933e-05,
+      "loss": 4.3381,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009632046898413e-05,
+      "loss": 4.3272,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001246099387893e-05,
+      "loss": 4.3365,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599286015187737e-05,
+      "loss": 4.3415,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598449058317059e-05,
+      "loss": 4.3444,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59761210144638e-05,
+      "loss": 4.3448,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596773506695328e-05,
+      "loss": 4.344,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595934911944276e-05,
+      "loss": 4.337,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595096317193224e-05,
+      "loss": 4.3446,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594257722442172e-05,
+      "loss": 4.3266,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59341912769112e-05,
+      "loss": 4.3385,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592580532940067e-05,
+      "loss": 4.3203,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591741938189015e-05,
+      "loss": 4.3363,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590903343437963e-05,
+      "loss": 4.3253,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590064748686911e-05,
+      "loss": 4.3347,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589226153935859e-05,
+      "loss": 4.3305,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588387559184808e-05,
+      "loss": 4.3234,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587550602314129e-05,
+      "loss": 4.3366,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586712007563077e-05,
+      "loss": 4.3283,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585873412812025e-05,
+      "loss": 4.326,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585034818060973e-05,
+      "loss": 4.3135,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584196223309921e-05,
+      "loss": 4.3173,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583357628558869e-05,
+      "loss": 4.3304,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582519033807817e-05,
+      "loss": 4.3252,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581680439056765e-05,
+      "loss": 4.3247,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808451200664585e-05,
+      "loss": 4.3111,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800065253154065e-05,
+      "loss": 4.3257,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791679305643545e-05,
+      "loss": 4.3098,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783293358133025e-05,
+      "loss": 4.3315,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577490741062251e-05,
+      "loss": 4.312,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576652146311199e-05,
+      "loss": 4.3203,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575813551560147e-05,
+      "loss": 4.3216,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574974956809095e-05,
+      "loss": 4.3078,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574139637818789e-05,
+      "loss": 4.3144,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573301043067737e-05,
+      "loss": 4.3205,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.3028,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3158,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570786896694954e-05,
+      "loss": 4.315,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3173,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.3055,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568271112441798e-05,
+      "loss": 4.3124,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.3,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665955608200674e-05,
+      "loss": 4.3133,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657569660690154e-05,
+      "loss": 4.324,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649183713179634e-05,
+      "loss": 4.3089,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640797765669114e-05,
+      "loss": 4.3226,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632411818158594e-05,
+      "loss": 4.3218,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624025870648074e-05,
+      "loss": 4.3217,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561565630194128e-05,
+      "loss": 4.2961,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560727035443076e-05,
+      "loss": 4.3168,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559888440692024e-05,
+      "loss": 4.315,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559049845940972e-05,
+      "loss": 4.304,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55821125118992e-05,
+      "loss": 4.3078,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557374294319242e-05,
+      "loss": 4.3094,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55653569956819e-05,
+      "loss": 4.3245,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555698742697511e-05,
+      "loss": 4.3089,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554860147946459e-05,
+      "loss": 4.2942,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554021553195407e-05,
+      "loss": 4.3072,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553182958444355e-05,
+      "loss": 4.3059,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552344363693303e-05,
+      "loss": 4.3112,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551505768942251e-05,
+      "loss": 4.3123,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550667174191198e-05,
+      "loss": 4.3009,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549828579440146e-05,
+      "loss": 4.3006,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548989984689094e-05,
+      "loss": 4.2938,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548151389938042e-05,
+      "loss": 4.2978,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54731279518699e-05,
+      "loss": 4.2994,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546474200435939e-05,
+      "loss": 4.3084,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545635605684887e-05,
+      "loss": 4.3199,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544797010933835e-05,
+      "loss": 4.3,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543958416182783e-05,
+      "loss": 4.2921,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431230971924765e-05,
+      "loss": 4.2959,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422845024414245e-05,
+      "loss": 4.2971,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414459076903725e-05,
+      "loss": 4.3052,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406073129393205e-05,
+      "loss": 4.2963,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397687181882685e-05,
+      "loss": 4.2997,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389301234372165e-05,
+      "loss": 4.2944,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380915286861645e-05,
+      "loss": 4.302,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372529339351125e-05,
+      "loss": 4.2772,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364143391840605e-05,
+      "loss": 4.292,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535577382313382e-05,
+      "loss": 4.2843,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534740425442703e-05,
+      "loss": 4.2974,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533901830691651e-05,
+      "loss": 4.2867,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533063235940599e-05,
+      "loss": 4.2961,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532224641189547e-05,
+      "loss": 4.2793,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531386046438495e-05,
+      "loss": 4.282,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530547451687443e-05,
+      "loss": 4.2871,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529708856936391e-05,
+      "loss": 4.2885,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528870262185339e-05,
+      "loss": 4.3035,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528031667434287e-05,
+      "loss": 4.2978,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527194710563608e-05,
+      "loss": 4.2916,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526356115812556e-05,
+      "loss": 4.2832,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525517521061504e-05,
+      "loss": 4.2881,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246805641908255e-05,
+      "loss": 4.2924,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238419694397735e-05,
+      "loss": 4.2836,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230033746887215e-05,
+      "loss": 4.2892,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221647799376694e-05,
+      "loss": 4.2842,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213278230669904e-05,
+      "loss": 4.2802,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5204892283159384e-05,
+      "loss": 4.2791,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196506335648863e-05,
+      "loss": 4.2804,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188120388138343e-05,
+      "loss": 4.2742,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179734440627817e-05,
+      "loss": 4.2851,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517136487192103e-05,
+      "loss": 4.2675,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516297892441051e-05,
+      "loss": 4.2951,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515459297689999e-05,
+      "loss": 4.2914,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514620702938947e-05,
+      "loss": 4.2932,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513782108187895e-05,
+      "loss": 4.2681,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512943513436843e-05,
+      "loss": 4.2833,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512104918685791e-05,
+      "loss": 4.2861,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511266323934739e-05,
+      "loss": 4.2842,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510427729183687e-05,
+      "loss": 4.2798,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509590772313008e-05,
+      "loss": 4.2805,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508752177561956e-05,
+      "loss": 4.2775,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507915220691277e-05,
+      "loss": 4.2874,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507076625940225e-05,
+      "loss": 4.2742,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506238031189173e-05,
+      "loss": 4.2807,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505399436438121e-05,
+      "loss": 4.2761,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50456084168707e-05,
+      "loss": 4.2702,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503722246936018e-05,
+      "loss": 4.2822,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502883652184966e-05,
+      "loss": 4.2785,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020466953142866e-05,
+      "loss": 4.2754,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012081005632346e-05,
+      "loss": 4.2681,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.2678,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.255432605743408,
+      "eval_runtime": 297.2024,
+      "eval_samples_per_second": 1283.943,
+      "eval_steps_per_second": 40.124,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.2695,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.2619,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2791,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2698,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2834,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2629,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.268,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2531,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2749,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.2738,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911466014309834e-05,
+      "loss": 4.2724,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903080066799314e-05,
+      "loss": 4.2711,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.2564,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886324550582003e-05,
+      "loss": 4.2667,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877938603071483e-05,
+      "loss": 4.2557,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486955265556096e-05,
+      "loss": 4.2637,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486116670805044e-05,
+      "loss": 4.2622,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485278076053992e-05,
+      "loss": 4.2584,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48443948130294e-05,
+      "loss": 4.2639,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483600886551888e-05,
+      "loss": 4.2767,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48276392968121e-05,
+      "loss": 4.2602,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481925334930158e-05,
+      "loss": 4.2676,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481086740179106e-05,
+      "loss": 4.258,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480248145428054e-05,
+      "loss": 4.2709,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479409550677002e-05,
+      "loss": 4.257,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2639,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.2559,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2549,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476056809553167e-05,
+      "loss": 4.2462,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2539,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2586,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2631,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.2616,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718654736782804e-05,
+      "loss": 4.2601,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710268789272283e-05,
+      "loss": 4.2567,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701882841761763e-05,
+      "loss": 4.2614,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469351327305497e-05,
+      "loss": 4.2514,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468512732554445e-05,
+      "loss": 4.253,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467674137803393e-05,
+      "loss": 4.2394,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466835543052341e-05,
+      "loss": 4.2525,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465996948301289e-05,
+      "loss": 4.2481,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651583535502366e-05,
+      "loss": 4.2549,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2511,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634811640481325e-05,
+      "loss": 4.2444,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626425692970805e-05,
+      "loss": 4.2572,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.2485,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.246,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460126785043925e-05,
+      "loss": 4.2358,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.234,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458452871302567e-05,
+      "loss": 4.2503,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457614276551515e-05,
+      "loss": 4.2454,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456775681800463e-05,
+      "loss": 4.2447,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455937087049411e-05,
+      "loss": 4.2358,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455098492298359e-05,
+      "loss": 4.2466,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454259897547307e-05,
+      "loss": 4.2317,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2501,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452584345925576e-05,
+      "loss": 4.2292,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451745751174524e-05,
+      "loss": 4.2442,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450907156423472e-05,
+      "loss": 4.2419,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500701995527935e-05,
+      "loss": 4.2358,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.2353,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483946479310624e-05,
+      "loss": 4.2426,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475560531800104e-05,
+      "loss": 4.2259,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467174584289584e-05,
+      "loss": 4.2396,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458788636779064e-05,
+      "loss": 4.2413,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450402689268544e-05,
+      "loss": 4.2394,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442016741758024e-05,
+      "loss": 4.2308,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433630794247504e-05,
+      "loss": 4.2354,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425244846736984e-05,
+      "loss": 4.2279,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441687527803019e-05,
+      "loss": 4.2331,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440848933051967e-05,
+      "loss": 4.246,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440010338300916e-05,
+      "loss": 4.2373,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439171743549864e-05,
+      "loss": 4.2447,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438333148798812e-05,
+      "loss": 4.2482,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437496191928133e-05,
+      "loss": 4.2472,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436657597177081e-05,
+      "loss": 4.2248,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435819002426029e-05,
+      "loss": 4.2365,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43498204555535e-05,
+      "loss": 4.2446,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434143450804298e-05,
+      "loss": 4.2307,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433304856053246e-05,
+      "loss": 4.2336,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432466261302194e-05,
+      "loss": 4.2328,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431627666551142e-05,
+      "loss": 4.2512,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.2373,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429952114929411e-05,
+      "loss": 4.2179,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429113520178359e-05,
+      "loss": 4.2324,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428274925427307e-05,
+      "loss": 4.2317,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427436330676255e-05,
+      "loss": 4.2377,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426599373805576e-05,
+      "loss": 4.2399,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425760779054524e-05,
+      "loss": 4.227,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424922184303472e-05,
+      "loss": 4.2257,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42408358955242e-05,
+      "loss": 4.2271,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232449948013675e-05,
+      "loss": 4.2218,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224064000503155e-05,
+      "loss": 4.2303,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2346,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207292105482115e-05,
+      "loss": 4.2465,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198906157971595e-05,
+      "loss": 4.2341,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419052021046108e-05,
+      "loss": 4.2188,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418213426295056e-05,
+      "loss": 4.2242,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417376469424377e-05,
+      "loss": 4.2232,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416537874673325e-05,
+      "loss": 4.2346,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415699279922273e-05,
+      "loss": 4.2288,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414860685171221e-05,
+      "loss": 4.2295,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414022090420169e-05,
+      "loss": 4.2222,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41318513354949e-05,
+      "loss": 4.232,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412346538798438e-05,
+      "loss": 4.2091,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411507944047386e-05,
+      "loss": 4.2239,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410669349296334e-05,
+      "loss": 4.2097,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409830754545282e-05,
+      "loss": 4.2299,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.2187,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408153565043178e-05,
+      "loss": 4.2239,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407314970292126e-05,
+      "loss": 4.2117,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064780134214475e-05,
+      "loss": 4.2151,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2207,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.2171,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039622291682915e-05,
+      "loss": 4.2328,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.2293,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402288315426933e-05,
+      "loss": 4.2234,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401451358556254e-05,
+      "loss": 4.2156,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400612763805202e-05,
+      "loss": 4.2166,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39977416905415e-05,
+      "loss": 4.2279,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398935574303098e-05,
+      "loss": 4.2081,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398096979552047e-05,
+      "loss": 4.226,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397258384800995e-05,
+      "loss": 4.21,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396419790049943e-05,
+      "loss": 4.2148,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395581195298891e-05,
+      "loss": 4.2156,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394742600547839e-05,
+      "loss": 4.2125,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393904005796786e-05,
+      "loss": 4.2032,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393065411045734e-05,
+      "loss": 4.2221,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392226816294682e-05,
+      "loss": 4.2034,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391389859424003e-05,
+      "loss": 4.2276,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390551264672951e-05,
+      "loss": 4.221,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389712669921899e-05,
+      "loss": 4.2301,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388874075170847e-05,
+      "loss": 4.198,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388035480419795e-05,
+      "loss": 4.2208,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387198523549117e-05,
+      "loss": 4.22,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863599287980647e-05,
+      "loss": 4.2192,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855213340470127e-05,
+      "loss": 4.2145,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846827392959606e-05,
+      "loss": 4.2166,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838457824252816e-05,
+      "loss": 4.2115,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830071876742296e-05,
+      "loss": 4.2205,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821685929231775e-05,
+      "loss": 4.2041,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813316360524985e-05,
+      "loss": 4.2163,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804930413014465e-05,
+      "loss": 4.2105,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796544465503944e-05,
+      "loss": 4.2074,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788158517993424e-05,
+      "loss": 4.2155,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377978894928664e-05,
+      "loss": 4.2159,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377140300177612e-05,
+      "loss": 4.2084,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37630170542656e-05,
+      "loss": 4.2,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375463110675508e-05,
+      "loss": 4.2074,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.19931697845459,
+      "eval_runtime": 308.7488,
+      "eval_samples_per_second": 1235.927,
+      "eval_steps_per_second": 38.624,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.2093,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373787559053777e-05,
+      "loss": 4.197,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372948964302725e-05,
+      "loss": 4.2118,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372110369551673e-05,
+      "loss": 4.2104,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371271774800621e-05,
+      "loss": 4.2178,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370433180049569e-05,
+      "loss": 4.2038,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369594585298517e-05,
+      "loss": 4.2024,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368755990547465e-05,
+      "loss": 4.1921,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367917395796413e-05,
+      "loss": 4.2089,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367078801045361e-05,
+      "loss": 4.2147,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366240206294309e-05,
+      "loss": 4.2088,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3654016115432576e-05,
+      "loss": 4.2085,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364563016792205e-05,
+      "loss": 4.1957,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363724422041153e-05,
+      "loss": 4.2039,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362885827290101e-05,
+      "loss": 4.1982,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362047232539049e-05,
+      "loss": 4.1998,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361208637787997e-05,
+      "loss": 4.1961,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360370043036945e-05,
+      "loss": 4.202,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359531448285893e-05,
+      "loss": 4.1958,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358694491415214e-05,
+      "loss": 4.2135,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357855896664162e-05,
+      "loss": 4.2035,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35701730191311e-05,
+      "loss": 4.2097,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356178707162058e-05,
+      "loss": 4.1947,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355341750291379e-05,
+      "loss": 4.2086,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354503155540327e-05,
+      "loss": 4.1935,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353664560789275e-05,
+      "loss": 4.2056,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352825966038223e-05,
+      "loss": 4.1986,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351987371287171e-05,
+      "loss": 4.1884,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351148776536119e-05,
+      "loss": 4.1849,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350310181785067e-05,
+      "loss": 4.1932,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349471587034015e-05,
+      "loss": 4.2007,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348636268043709e-05,
+      "loss": 4.2014,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347797673292657e-05,
+      "loss": 4.2013,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.2034,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.1968,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345281889039501e-05,
+      "loss": 4.2006,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344444932168823e-05,
+      "loss": 4.1938,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343606337417771e-05,
+      "loss": 4.1932,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342767742666719e-05,
+      "loss": 4.1797,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341929147915667e-05,
+      "loss": 4.1946,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341090553164615e-05,
+      "loss": 4.188,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340251958413563e-05,
+      "loss": 4.1967,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339413363662511e-05,
+      "loss": 4.1916,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338574768911458e-05,
+      "loss": 4.1878,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.1984,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369008551701005e-05,
+      "loss": 4.1871,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360622604190485e-05,
+      "loss": 4.188,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352236656679965e-05,
+      "loss": 4.1788,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343850709169445e-05,
+      "loss": 4.1756,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333546476165893e-05,
+      "loss": 4.191,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332709519295214e-05,
+      "loss": 4.1891,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331870924544162e-05,
+      "loss": 4.1858,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1803,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.1895,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.1709,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285165455399534e-05,
+      "loss": 4.1953,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3276779507889013e-05,
+      "loss": 4.1742,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268393560378493e-05,
+      "loss": 4.1868,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.183,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.1812,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324325209665067e-05,
+      "loss": 4.1767,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323486614914015e-05,
+      "loss": 4.1847,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322648020162963e-05,
+      "loss": 4.1756,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321811063292284e-05,
+      "loss": 4.1798,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320972468541232e-05,
+      "loss": 4.1841,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32013387379018e-05,
+      "loss": 4.1827,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319295279039128e-05,
+      "loss": 4.1719,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1839,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.1672,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1766,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.1899,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.1841,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314266986293562e-05,
+      "loss": 4.1875,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31342839154251e-05,
+      "loss": 4.1882,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312589796791458e-05,
+      "loss": 4.1952,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311752839920779e-05,
+      "loss": 4.1727,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310914245169727e-05,
+      "loss": 4.1781,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1887,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309238693547996e-05,
+      "loss": 4.1722,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308400098796944e-05,
+      "loss": 4.1832,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307561504045892e-05,
+      "loss": 4.1814,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30672290929484e-05,
+      "loss": 4.1893,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.184,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305047357673109e-05,
+      "loss": 4.1666,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042087629220577e-05,
+      "loss": 4.1773,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3033701681710057e-05,
+      "loss": 4.1767,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025315734199536e-05,
+      "loss": 4.1835,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016946165492746e-05,
+      "loss": 4.1822,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1768,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1753,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2991788322961185e-05,
+      "loss": 4.1698,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983418754254395e-05,
+      "loss": 4.1662,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975032806743874e-05,
+      "loss": 4.1722,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966646859233354e-05,
+      "loss": 4.1832,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.1944,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949891343016043e-05,
+      "loss": 4.1834,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294150539550553e-05,
+      "loss": 4.1608,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293311944799501e-05,
+      "loss": 4.1739,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1716,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291634755297397e-05,
+      "loss": 4.1808,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290796160546345e-05,
+      "loss": 4.1734,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.175,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.1723,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288283652053935e-05,
+      "loss": 4.1742,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287445057302883e-05,
+      "loss": 4.1623,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286606462551831e-05,
+      "loss": 4.1709,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285767867800779e-05,
+      "loss": 4.1558,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849309109301e-05,
+      "loss": 4.1792,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284092316179048e-05,
+      "loss": 4.1661,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2832537214279964e-05,
+      "loss": 4.168,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824151266769444e-05,
+      "loss": 4.1567,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2815765319258924e-05,
+      "loss": 4.1677,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807379371748404e-05,
+      "loss": 4.1676,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2798993424237884e-05,
+      "loss": 4.1658,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279062385553109e-05,
+      "loss": 4.1811,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278223790802057e-05,
+      "loss": 4.1776,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277386833931378e-05,
+      "loss": 4.1728,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276548239180326e-05,
+      "loss": 4.1673,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275709644429274e-05,
+      "loss": 4.1601,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274871049678222e-05,
+      "loss": 4.1765,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1586,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1732,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1563,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271516670674014e-05,
+      "loss": 4.1692,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270679713803336e-05,
+      "loss": 4.1581,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269841119052284e-05,
+      "loss": 4.1649,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269002524301231e-05,
+      "loss": 4.1579,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268163929550179e-05,
+      "loss": 4.1636,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267325334799127e-05,
+      "loss": 4.1578,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266486740048075e-05,
+      "loss": 4.1744,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265649783177396e-05,
+      "loss": 4.175,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264811188426344e-05,
+      "loss": 4.1768,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263972593675292e-05,
+      "loss": 4.1483,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.1722,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622954041731886e-05,
+      "loss": 4.1691,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614568094221366e-05,
+      "loss": 4.1683,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1631,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597796199200326e-05,
+      "loss": 4.1681,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1647,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581057061786744e-05,
+      "loss": 4.172,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2572671114276224e-05,
+      "loss": 4.1532,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564285166765704e-05,
+      "loss": 4.1698,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2555899219255184e-05,
+      "loss": 4.1593,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2547513271744664e-05,
+      "loss": 4.1593,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539127324234144e-05,
+      "loss": 4.1659,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2530741376723624e-05,
+      "loss": 4.1612,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252238818682057e-05,
+      "loss": 4.1659,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251400223931005e-05,
+      "loss": 4.1503,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250561629179953e-05,
+      "loss": 4.1611,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.158422946929932,
+      "eval_runtime": 296.3021,
+      "eval_samples_per_second": 1287.844,
+      "eval_steps_per_second": 40.246,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249723034428901e-05,
+      "loss": 4.1641,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248884439677849e-05,
+      "loss": 4.1485,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248045844926797e-05,
+      "loss": 4.162,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247207250175745e-05,
+      "loss": 4.1633,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246368655424693e-05,
+      "loss": 4.1688,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245531698554014e-05,
+      "loss": 4.154,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244693103802962e-05,
+      "loss": 4.1498,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24385450905191e-05,
+      "loss": 4.1508,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243017552181231e-05,
+      "loss": 4.1547,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2421789574301793e-05,
+      "loss": 4.1668,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241340362679127e-05,
+      "loss": 4.1597,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240501767928075e-05,
+      "loss": 4.1597,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239663173177023e-05,
+      "loss": 4.1521,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238824578425971e-05,
+      "loss": 4.1535,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237985983674919e-05,
+      "loss": 4.1526,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237147388923867e-05,
+      "loss": 4.1506,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236310432053188e-05,
+      "loss": 4.1504,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235471837302136e-05,
+      "loss": 4.1507,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234633242551084e-05,
+      "loss": 4.1519,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2337946478000315e-05,
+      "loss": 4.1638,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2329560530489795e-05,
+      "loss": 4.1574,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2321174582979275e-05,
+      "loss": 4.1631,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231280501427249e-05,
+      "loss": 4.1453,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230441906676197e-05,
+      "loss": 4.1627,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229603311925145e-05,
+      "loss": 4.1458,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228764717174093e-05,
+      "loss": 4.1585,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227926122423041e-05,
+      "loss": 4.1479,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227087527671989e-05,
+      "loss": 4.1403,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226248932920937e-05,
+      "loss": 4.1444,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225410338169885e-05,
+      "loss": 4.1466,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224573381299206e-05,
+      "loss": 4.152,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223734786548154e-05,
+      "loss": 4.1558,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2228994675578485e-05,
+      "loss": 4.1537,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2220608728067965e-05,
+      "loss": 4.1614,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212222780557445e-05,
+      "loss": 4.1431,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2203836833046925e-05,
+      "loss": 4.1566,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219546726434014e-05,
+      "loss": 4.1516,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218708131682962e-05,
+      "loss": 4.1421,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2178695369319094e-05,
+      "loss": 4.1362,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2170309421808574e-05,
+      "loss": 4.1426,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2161923474298054e-05,
+      "loss": 4.1484,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2153537526787534e-05,
+      "loss": 4.1511,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2145151579277014e-05,
+      "loss": 4.144,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2136765631766494e-05,
+      "loss": 4.1396,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128379684255973e-05,
+      "loss": 4.1576,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212001011554918e-05,
+      "loss": 4.1433,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211162416803866e-05,
+      "loss": 4.1421,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210323822052815e-05,
+      "loss": 4.1335,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209485227301763e-05,
+      "loss": 4.13,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208646632550711e-05,
+      "loss": 4.1464,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207808037799659e-05,
+      "loss": 4.1438,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206969443048607e-05,
+      "loss": 4.142,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1328,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205293891426876e-05,
+      "loss": 4.1479,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204456934556197e-05,
+      "loss": 4.1241,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1513,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.1307,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.1358,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1428,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1332,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199427003930258e-05,
+      "loss": 4.1346,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198588409179206e-05,
+      "loss": 4.1453,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197749814428154e-05,
+      "loss": 4.1303,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196911219677102e-05,
+      "loss": 4.1349,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19607262492605e-05,
+      "loss": 4.1412,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195234030174998e-05,
+      "loss": 4.1414,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194397073304319e-05,
+      "loss": 4.1303,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19356011643364e-05,
+      "loss": 4.14,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192721521682588e-05,
+      "loss": 4.122,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191882926931536e-05,
+      "loss": 4.1338,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191044332180484e-05,
+      "loss": 4.1448,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190205737429432e-05,
+      "loss": 4.1394,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18936714267838e-05,
+      "loss": 4.1449,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188528547927328e-05,
+      "loss": 4.1412,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187689953176276e-05,
+      "loss": 4.1572,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186851358425224e-05,
+      "loss": 4.1277,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860144015545456e-05,
+      "loss": 4.1415,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1851774446838665e-05,
+      "loss": 4.1464,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1843388499328145e-05,
+      "loss": 4.1254,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835002551817625e-05,
+      "loss": 4.1404,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18266166043071e-05,
+      "loss": 4.1395,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181823065679658e-05,
+      "loss": 4.1492,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180984470928606e-05,
+      "loss": 4.1403,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180145876177554e-05,
+      "loss": 4.1245,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793072814265025e-05,
+      "loss": 4.1312,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1784703245558234e-05,
+      "loss": 4.1328,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776317298047714e-05,
+      "loss": 4.1418,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767931350537194e-05,
+      "loss": 4.1415,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.17595617818304e-05,
+      "loss": 4.1383,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175117583431988e-05,
+      "loss": 4.1241,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174278988680936e-05,
+      "loss": 4.1297,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173440393929884e-05,
+      "loss": 4.1267,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172601799178832e-05,
+      "loss": 4.1323,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17176320442778e-05,
+      "loss": 4.1421,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170926247557101e-05,
+      "loss": 4.1487,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170087652806049e-05,
+      "loss": 4.143,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169249058054998e-05,
+      "loss": 4.1233,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168410463303946e-05,
+      "loss": 4.1298,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167571868552894e-05,
+      "loss": 4.1295,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166733273801842e-05,
+      "loss": 4.1372,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16589467905079e-05,
+      "loss": 4.1321,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165056084299738e-05,
+      "loss": 4.1395,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164219127429059e-05,
+      "loss": 4.1317,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163380532678007e-05,
+      "loss": 4.1258,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162541937926955e-05,
+      "loss": 4.1249,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161703343175903e-05,
+      "loss": 4.1284,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160866386305224e-05,
+      "loss": 4.1172,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160027791554172e-05,
+      "loss": 4.1374,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1591891968031197e-05,
+      "loss": 4.1268,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1583506020520677e-05,
+      "loss": 4.1295,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157513645181389e-05,
+      "loss": 4.1185,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.1235,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155838093559658e-05,
+      "loss": 4.1303,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154999498808606e-05,
+      "loss": 4.1232,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154160904057554e-05,
+      "loss": 4.1427,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153323947186875e-05,
+      "loss": 4.1337,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152486990316196e-05,
+      "loss": 4.1332,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151648395565144e-05,
+      "loss": 4.1316,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150809800814092e-05,
+      "loss": 4.1147,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14997120606304e-05,
+      "loss": 4.138,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149132611311988e-05,
+      "loss": 4.1246,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1482940165609366e-05,
+      "loss": 4.1312,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1474554218098846e-05,
+      "loss": 4.1176,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.1272,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1457798701881535e-05,
+      "loss": 4.1185,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1449412754371015e-05,
+      "loss": 4.1247,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1441026806860495e-05,
+      "loss": 4.1168,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1432640859349975e-05,
+      "loss": 4.1291,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1424254911839455e-05,
+      "loss": 4.1161,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1415868964328935e-05,
+      "loss": 4.1358,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140748301681841e-05,
+      "loss": 4.1293,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139909706930789e-05,
+      "loss": 4.1371,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139071112179737e-05,
+      "loss": 4.114,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138232517428685e-05,
+      "loss": 4.1341,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1373939226776335e-05,
+      "loss": 4.1287,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1365553279265815e-05,
+      "loss": 4.1291,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357183710559024e-05,
+      "loss": 4.1266,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1348797763048504e-05,
+      "loss": 4.1293,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340411815537984e-05,
+      "loss": 4.1212,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1332025868027464e-05,
+      "loss": 4.1409,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132367267812441e-05,
+      "loss": 4.1115,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131528673061388e-05,
+      "loss": 4.1302,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1215,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129851483559284e-05,
+      "loss": 4.1221,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129012888808232e-05,
+      "loss": 4.1275,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12817429405718e-05,
+      "loss": 4.1224,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127335699306129e-05,
+      "loss": 4.1297,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12649874243545e-05,
+      "loss": 4.1105,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1256617855647707e-05,
+      "loss": 4.124,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.128399848937988,
+      "eval_runtime": 294.7977,
+      "eval_samples_per_second": 1294.416,
+      "eval_steps_per_second": 40.451,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248231908137186e-05,
+      "loss": 4.1246,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1239845960626666e-05,
+      "loss": 4.1088,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1231460013116146e-05,
+      "loss": 4.1251,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1223074065605626e-05,
+      "loss": 4.1203,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1214688118095106e-05,
+      "loss": 4.1294,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206302170584586e-05,
+      "loss": 4.1211,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1197916223074066e-05,
+      "loss": 4.114,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1189530275563546e-05,
+      "loss": 4.113,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1181144328053026e-05,
+      "loss": 4.1178,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1172758380542506e-05,
+      "loss": 4.1276,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1164372433031986e-05,
+      "loss": 4.1237,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115598648552147e-05,
+      "loss": 4.1202,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114760053801095e-05,
+      "loss": 4.1156,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1139214590500426e-05,
+      "loss": 4.1157,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1130828642989906e-05,
+      "loss": 4.1124,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122442695479386e-05,
+      "loss": 4.1112,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1114073126772595e-05,
+      "loss": 4.1158,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1105687179262075e-05,
+      "loss": 4.1122,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1097301231751555e-05,
+      "loss": 4.1168,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1088915284241035e-05,
+      "loss": 4.1214,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1080545715534244e-05,
+      "loss": 4.1232,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072159768023724e-05,
+      "loss": 4.1242,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106377382051321e-05,
+      "loss": 4.1109,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105538787300269e-05,
+      "loss": 4.1226,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104700192549217e-05,
+      "loss": 4.1092,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103861597798165e-05,
+      "loss": 4.1261,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103023003047113e-05,
+      "loss": 4.108,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102184408296061e-05,
+      "loss": 4.1059,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101345813545009e-05,
+      "loss": 4.1087,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100507218793957e-05,
+      "loss": 4.1131,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099668624042905e-05,
+      "loss": 4.1117,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098831667172226e-05,
+      "loss": 4.1197,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097994710301547e-05,
+      "loss": 4.1189,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097156115550495e-05,
+      "loss": 4.1202,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096317520799443e-05,
+      "loss": 4.1114,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095478926048391e-05,
+      "loss": 4.1188,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094643607058085e-05,
+      "loss": 4.1174,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093805012307033e-05,
+      "loss": 4.1014,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092966417555981e-05,
+      "loss": 4.1037,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092127822804929e-05,
+      "loss": 4.1062,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091289228053877e-05,
+      "loss": 4.1093,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.090450633302825e-05,
+      "loss": 4.1162,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089612038551773e-05,
+      "loss": 4.1078,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088773443800721e-05,
+      "loss": 4.1048,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087934849049669e-05,
+      "loss": 4.1188,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087096254298617e-05,
+      "loss": 4.1067,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086257659547565e-05,
+      "loss": 4.111,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085420702676886e-05,
+      "loss": 4.0943,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084582107925834e-05,
+      "loss": 4.09,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083743513174783e-05,
+      "loss": 4.1174,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082904918423731e-05,
+      "loss": 4.1052,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082067961553052e-05,
+      "loss": 4.1103,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081229366802e-05,
+      "loss": 4.1022,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080390772050948e-05,
+      "loss": 4.1048,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079552177299896e-05,
+      "loss": 4.0911,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078713582548843e-05,
+      "loss": 4.1101,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077874987797791e-05,
+      "loss": 4.0997,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077036393046739e-05,
+      "loss": 4.1015,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076197798295687e-05,
+      "loss": 4.1128,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075359203544635e-05,
+      "loss": 4.0981,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745222466739566e-05,
+      "loss": 4.0966,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0736836519229046e-05,
+      "loss": 4.111,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0728450571718526e-05,
+      "loss": 4.0948,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720081003011735e-05,
+      "loss": 4.0956,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0711695055501215e-05,
+      "loss": 4.1078,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703309107990695e-05,
+      "loss": 4.1046,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0694923160480175e-05,
+      "loss": 4.0916,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0686537212969655e-05,
+      "loss": 4.1077,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0678151265459135e-05,
+      "loss": 4.0875,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0669781696752344e-05,
+      "loss": 4.0971,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0661395749241824e-05,
+      "loss": 4.109,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653009801731304e-05,
+      "loss": 4.1067,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0644623854220784e-05,
+      "loss": 4.1059,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636237906710264e-05,
+      "loss": 4.106,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062785195919975e-05,
+      "loss": 4.1235,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061946601168923e-05,
+      "loss": 4.0965,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061109644298244e-05,
+      "loss": 4.1051,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060271049547192e-05,
+      "loss": 4.1125,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05943245479614e-05,
+      "loss": 4.0938,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058593860045088e-05,
+      "loss": 4.1032,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057756903174409e-05,
+      "loss": 4.1084,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056918308423357e-05,
+      "loss": 4.1114,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056079713672305e-05,
+      "loss": 4.1081,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055241118921253e-05,
+      "loss": 4.0925,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054402524170201e-05,
+      "loss": 4.0911,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053563929419149e-05,
+      "loss": 4.1016,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052725334668097e-05,
+      "loss": 4.1082,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051886739917045e-05,
+      "loss": 4.1068,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510497830463664e-05,
+      "loss": 4.1048,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502111882953144e-05,
+      "loss": 4.0863,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049372593544262e-05,
+      "loss": 4.0966,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04853399879321e-05,
+      "loss": 4.0952,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047695404042158e-05,
+      "loss": 4.1007,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046856809291106e-05,
+      "loss": 4.1046,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046018214540054e-05,
+      "loss": 4.1167,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451812576693746e-05,
+      "loss": 4.1053,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443426629183226e-05,
+      "loss": 4.0928,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0435040681672706e-05,
+      "loss": 4.0937,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0426654734162186e-05,
+      "loss": 4.097,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418268786651666e-05,
+      "loss": 4.1062,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040988283914115e-05,
+      "loss": 4.0977,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040151327043436e-05,
+      "loss": 4.1078,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039312732292384e-05,
+      "loss": 4.1006,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038474137541332e-05,
+      "loss": 4.0874,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03763554279028e-05,
+      "loss": 4.0969,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036796948039228e-05,
+      "loss": 4.0925,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035958353288176e-05,
+      "loss": 4.0887,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035119758537124e-05,
+      "loss": 4.0992,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034281163786072e-05,
+      "loss": 4.0964,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033444206915393e-05,
+      "loss": 4.0928,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032605612164341e-05,
+      "loss": 4.0891,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031767017413289e-05,
+      "loss": 4.0868,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030928422662237e-05,
+      "loss": 4.0968,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030089827911185e-05,
+      "loss": 4.0882,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029252871040507e-05,
+      "loss": 4.1115,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284142762894547e-05,
+      "loss": 4.1027,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0275756815384027e-05,
+      "loss": 4.1009,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267387246677236e-05,
+      "loss": 4.0974,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259001299166716e-05,
+      "loss": 4.0811,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250615351656196e-05,
+      "loss": 4.1001,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242229404145675e-05,
+      "loss": 4.0982,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233843456635155e-05,
+      "loss": 4.0956,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225473887928365e-05,
+      "loss": 4.0849,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217087940417845e-05,
+      "loss": 4.0982,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0208701992907324e-05,
+      "loss": 4.0866,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200316045396804e-05,
+      "loss": 4.0921,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191930097886284e-05,
+      "loss": 4.0825,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01835605291795e-05,
+      "loss": 4.0985,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017517458166898e-05,
+      "loss": 4.0792,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016678863415845e-05,
+      "loss": 4.1058,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015840268664793e-05,
+      "loss": 4.0946,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015003311794115e-05,
+      "loss": 4.1038,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014164717043063e-05,
+      "loss": 4.0826,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01332612229201e-05,
+      "loss": 4.1012,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012487527540958e-05,
+      "loss": 4.096,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011648932789906e-05,
+      "loss": 4.0997,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010810338038854e-05,
+      "loss": 4.0949,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009971743287803e-05,
+      "loss": 4.0924,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009133148536751e-05,
+      "loss": 4.0924,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008294553785699e-05,
+      "loss": 4.1064,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007455959034647e-05,
+      "loss": 4.0799,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006619002163968e-05,
+      "loss": 4.0974,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005780407412916e-05,
+      "loss": 4.0934,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004941812661864e-05,
+      "loss": 4.0876,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004103217910812e-05,
+      "loss": 4.0949,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00326462315976e-05,
+      "loss": 4.0874,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002426028408708e-05,
+      "loss": 4.1,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001589071538029e-05,
+      "loss": 4.077,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000750476786977e-05,
+      "loss": 4.0944,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.104121208190918,
+      "eval_runtime": 295.5435,
+      "eval_samples_per_second": 1291.15,
+      "eval_steps_per_second": 40.349,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999911882035925e-05,
+      "loss": 4.0907,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999073287284873e-05,
+      "loss": 4.0807,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9982346925338207e-05,
+      "loss": 4.0888,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997396097782769e-05,
+      "loss": 4.0893,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9965575030317167e-05,
+      "loss": 4.1019,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995720546161038e-05,
+      "loss": 4.0855,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994881951409986e-05,
+      "loss": 4.0858,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994043356658934e-05,
+      "loss": 4.083,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9932047619078815e-05,
+      "loss": 4.0839,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9923661671568295e-05,
+      "loss": 4.0975,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9915275724057775e-05,
+      "loss": 4.0882,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9906906155350984e-05,
+      "loss": 4.0925,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9898520207840464e-05,
+      "loss": 4.0828,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9890134260329944e-05,
+      "loss": 4.0855,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988174831281943e-05,
+      "loss": 4.0828,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987336236530891e-05,
+      "loss": 4.0794,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986497641779839e-05,
+      "loss": 4.0826,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98566068490916e-05,
+      "loss": 4.0838,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984822090158108e-05,
+      "loss": 4.0842,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983983495407056e-05,
+      "loss": 4.088,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983144900656004e-05,
+      "loss": 4.0945,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982306305904952e-05,
+      "loss": 4.0957,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981469349034273e-05,
+      "loss": 4.0779,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980632392163594e-05,
+      "loss": 4.0902,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979793797412542e-05,
+      "loss": 4.0815,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97895520266149e-05,
+      "loss": 4.0932,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781166079104385e-05,
+      "loss": 4.0808,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9772780131593865e-05,
+      "loss": 4.0761,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9764394184083345e-05,
+      "loss": 4.0791,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9756008236572825e-05,
+      "loss": 4.0767,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9747622289062305e-05,
+      "loss": 4.0759,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9739236341551785e-05,
+      "loss": 4.0909,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9730850394041265e-05,
+      "loss": 4.0843,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9722464446530744e-05,
+      "loss": 4.0896,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9714094877823954e-05,
+      "loss": 4.0814,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705708930313434e-05,
+      "loss": 4.089,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969733936160664e-05,
+      "loss": 4.0879,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968895341409612e-05,
+      "loss": 4.0706,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96805674665856e-05,
+      "loss": 4.0754,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967218151907508e-05,
+      "loss": 4.0735,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966379557156457e-05,
+      "loss": 4.0758,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965540962405405e-05,
+      "loss": 4.0857,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964702367654353e-05,
+      "loss": 4.0767,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9638637729033e-05,
+      "loss": 4.0755,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963025178152248e-05,
+      "loss": 4.087,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962186583401196e-05,
+      "loss": 4.0743,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961347988650144e-05,
+      "loss": 4.0833,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960509393899092e-05,
+      "loss": 4.0667,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959672437028413e-05,
+      "loss": 4.0566,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958835480157735e-05,
+      "loss": 4.0879,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957996885406682e-05,
+      "loss": 4.0745,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957158290655631e-05,
+      "loss": 4.0812,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956319695904579e-05,
+      "loss": 4.0722,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955481101153527e-05,
+      "loss": 4.0693,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954642506402475e-05,
+      "loss": 4.0669,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9538055495317956e-05,
+      "loss": 4.0776,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9529669547807436e-05,
+      "loss": 4.0667,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9521283600296916e-05,
+      "loss": 4.0736,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9512897652786396e-05,
+      "loss": 4.0836,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9504511705275876e-05,
+      "loss": 4.0653,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9496142136569085e-05,
+      "loss": 4.0658,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9487756189058565e-05,
+      "loss": 4.0837,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9479370241548045e-05,
+      "loss": 4.0628,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9470984294037525e-05,
+      "loss": 4.0657,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9462598346527005e-05,
+      "loss": 4.0806,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945422877782022e-05,
+      "loss": 4.072,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94458428303097e-05,
+      "loss": 4.0644,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943745688279918e-05,
+      "loss": 4.078,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942908731409239e-05,
+      "loss": 4.0592,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942070136658187e-05,
+      "loss": 4.065,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941231541907135e-05,
+      "loss": 4.0811,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940392947156083e-05,
+      "loss": 4.0755,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939554352405031e-05,
+      "loss": 4.0763,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938715757653979e-05,
+      "loss": 4.0775,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937877162902927e-05,
+      "loss": 4.0955,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937038568151875e-05,
+      "loss": 4.0637,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936199973400823e-05,
+      "loss": 4.0771,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0807,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345260596594654e-05,
+      "loss": 4.069,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9336874649084134e-05,
+      "loss": 4.0685,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9328488701573614e-05,
+      "loss": 4.0834,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9320102754063094e-05,
+      "loss": 4.0792,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9311716806552574e-05,
+      "loss": 4.0829,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9303330859042054e-05,
+      "loss": 4.061,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9294944911531534e-05,
+      "loss": 4.0589,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928655896402101e-05,
+      "loss": 4.0733,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927818939531422e-05,
+      "loss": 4.0805,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92698034478037e-05,
+      "loss": 4.0769,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926141750029318e-05,
+      "loss": 4.076,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925304793158639e-05,
+      "loss": 4.0574,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924466198407588e-05,
+      "loss": 4.0684,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923627603656536e-05,
+      "loss": 4.0662,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922789008905484e-05,
+      "loss": 4.0667,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921950414154431e-05,
+      "loss": 4.0773,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921111819403379e-05,
+      "loss": 4.0844,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920273224652327e-05,
+      "loss": 4.077,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919434629901275e-05,
+      "loss": 4.0674,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918596035150223e-05,
+      "loss": 4.0617,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917759078279544e-05,
+      "loss": 4.0696,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916920483528492e-05,
+      "loss": 4.078,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91608188877744e-05,
+      "loss": 4.0697,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915243294026388e-05,
+      "loss": 4.0791,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914404699275336e-05,
+      "loss": 4.0712,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913566104524285e-05,
+      "loss": 4.0622,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912727509773233e-05,
+      "loss": 4.0676,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911888915022181e-05,
+      "loss": 4.0623,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9110535960318745e-05,
+      "loss": 4.0601,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9102150012808225e-05,
+      "loss": 4.0711,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9093780444101435e-05,
+      "loss": 4.0701,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9085394496590914e-05,
+      "loss": 4.0659,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9077008549080394e-05,
+      "loss": 4.059,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9068622601569874e-05,
+      "loss": 4.0565,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9060236654059354e-05,
+      "loss": 4.0697,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9051850706548834e-05,
+      "loss": 4.0595,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9043464759038314e-05,
+      "loss": 4.0832,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903509519033153e-05,
+      "loss": 4.0766,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902670924282101e-05,
+      "loss": 4.072,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901832329531049e-05,
+      "loss": 4.0663,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900993734779997e-05,
+      "loss": 4.0586,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900155140028945e-05,
+      "loss": 4.0702,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899316545277893e-05,
+      "loss": 4.0684,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898477950526841e-05,
+      "loss": 4.067,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897639355775789e-05,
+      "loss": 4.0583,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896800761024737e-05,
+      "loss": 4.0706,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895962166273684e-05,
+      "loss": 4.0589,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895123571522632e-05,
+      "loss": 4.0614,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89428497677158e-05,
+      "loss": 4.0533,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.893448019900902e-05,
+      "loss": 4.0732,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89260942514985e-05,
+      "loss": 4.0535,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891770830398798e-05,
+      "loss": 4.0732,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.0693,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8900952787770674e-05,
+      "loss": 4.0761,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889256684026015e-05,
+      "loss": 4.057,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888418089274963e-05,
+      "loss": 4.0757,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887579494523911e-05,
+      "loss": 4.0647,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886740899772859e-05,
+      "loss": 4.0718,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885902305021807e-05,
+      "loss": 4.0707,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885063710270755e-05,
+      "loss": 4.0608,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8842267534000757e-05,
+      "loss": 4.0662,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8833881586490236e-05,
+      "loss": 4.0769,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8825495638979716e-05,
+      "loss": 4.0515,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88171096914692e-05,
+      "loss": 4.0664,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880874012276241e-05,
+      "loss": 4.0717,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880035417525189e-05,
+      "loss": 4.0634,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.87919846065451e-05,
+      "loss": 4.0598,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878359865903458e-05,
+      "loss": 4.0617,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877521271152406e-05,
+      "loss": 4.0716,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876682676401354e-05,
+      "loss": 4.0528,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875844081650302e-05,
+      "loss": 4.0657,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0843682289123535,
+      "eval_runtime": 295.1155,
+      "eval_samples_per_second": 1293.022,
+      "eval_steps_per_second": 40.408,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87500548689925e-05,
+      "loss": 4.0597,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874166892148198e-05,
+      "loss": 4.0553,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873328297397146e-05,
+      "loss": 4.0581,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872489702646094e-05,
+      "loss": 4.0642,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871651107895042e-05,
+      "loss": 4.076,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87081251314399e-05,
+      "loss": 4.0575,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869973918392939e-05,
+      "loss": 4.0581,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869135323641886e-05,
+      "loss": 4.0595,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868296728890834e-05,
+      "loss": 4.0519,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867458134139782e-05,
+      "loss": 4.0737,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86661953938873e-05,
+      "loss": 4.0595,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865780944637678e-05,
+      "loss": 4.0623,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864942349886626e-05,
+      "loss": 4.0645,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864103755135574e-05,
+      "loss": 4.054,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863265160384522e-05,
+      "loss": 4.0541,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86242656563347e-05,
+      "loss": 4.0529,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861587970882418e-05,
+      "loss": 4.0589,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860751014011739e-05,
+      "loss": 4.0604,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859912419260687e-05,
+      "loss": 4.0548,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590738245096356e-05,
+      "loss": 4.0601,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8582352297585836e-05,
+      "loss": 4.0698,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8573966350075316e-05,
+      "loss": 4.0699,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565596781368525e-05,
+      "loss": 4.059,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8557227212661734e-05,
+      "loss": 4.0552,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8548841265151214e-05,
+      "loss": 4.0565,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8540455317640694e-05,
+      "loss": 4.0663,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8532069370130174e-05,
+      "loss": 4.0559,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8523683422619654e-05,
+      "loss": 4.0513,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8515297475109134e-05,
+      "loss": 4.0532,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8506911527598614e-05,
+      "loss": 4.0495,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498525580088094e-05,
+      "loss": 4.0509,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490139632577574e-05,
+      "loss": 4.0673,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8481753685067054e-05,
+      "loss": 4.0611,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473367737556534e-05,
+      "loss": 4.0639,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846499816884975e-05,
+      "loss": 4.0573,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845661222133922e-05,
+      "loss": 4.0561,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84482262738287e-05,
+      "loss": 4.0665,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843984032631818e-05,
+      "loss": 4.0492,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843145437880766e-05,
+      "loss": 4.052,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842306843129714e-05,
+      "loss": 4.0462,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841468248378662e-05,
+      "loss": 4.05,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84062965362761e-05,
+      "loss": 4.0595,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839792696756931e-05,
+      "loss": 4.0519,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838954102005879e-05,
+      "loss": 4.0517,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838115507254827e-05,
+      "loss": 4.0619,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837276912503776e-05,
+      "loss": 4.0505,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836438317752724e-05,
+      "loss": 4.0552,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835601360882045e-05,
+      "loss": 4.0462,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834762766130993e-05,
+      "loss": 4.0246,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833924171379941e-05,
+      "loss": 4.0672,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8330872145092616e-05,
+      "loss": 4.0479,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8322486197582096e-05,
+      "loss": 4.061,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314100250071576e-05,
+      "loss": 4.0436,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8305714302561056e-05,
+      "loss": 4.0486,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8297328355050536e-05,
+      "loss": 4.0433,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8288958786343745e-05,
+      "loss": 4.0485,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8280572838833225e-05,
+      "loss": 4.0442,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827218689132271e-05,
+      "loss": 4.0484,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826380094381219e-05,
+      "loss": 4.0617,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825541499630167e-05,
+      "loss": 4.0408,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824702904879115e-05,
+      "loss": 4.0412,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823864310128063e-05,
+      "loss": 4.0583,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8230257153770105e-05,
+      "loss": 4.0418,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822190396386705e-05,
+      "loss": 4.0393,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821351801635653e-05,
+      "loss": 4.0546,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820513206884601e-05,
+      "loss": 4.0482,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819674612133549e-05,
+      "loss": 4.0405,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818836017382497e-05,
+      "loss": 4.052,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817999060511818e-05,
+      "loss": 4.0356,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171604657607666e-05,
+      "loss": 4.0446,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163218710097146e-05,
+      "loss": 4.0556,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154832762586626e-05,
+      "loss": 4.0513,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146446815076106e-05,
+      "loss": 4.0536,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813806086756558e-05,
+      "loss": 4.0506,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812967492005506e-05,
+      "loss": 4.0695,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812128897254454e-05,
+      "loss": 4.0393,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112919403837755e-05,
+      "loss": 4.0585,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810453345632723e-05,
+      "loss": 4.0502,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809614750881671e-05,
+      "loss": 4.0482,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808776156130619e-05,
+      "loss": 4.0446,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8079391992599403e-05,
+      "loss": 4.0591,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071006045088883e-05,
+      "loss": 4.0551,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806262009757836e-05,
+      "loss": 4.0604,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805423415006784e-05,
+      "loss": 4.0371,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804584820255732e-05,
+      "loss": 4.0399,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80374622550468e-05,
+      "loss": 4.048,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802907630753628e-05,
+      "loss": 4.0533,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802070673882949e-05,
+      "loss": 4.0478,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801232079131897e-05,
+      "loss": 4.056,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800393484380845e-05,
+      "loss": 4.0375,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799554889629793e-05,
+      "loss": 4.0455,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798717932759114e-05,
+      "loss": 4.0435,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797879338008062e-05,
+      "loss": 4.0401,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79704074325701e-05,
+      "loss": 4.0537,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796202148505959e-05,
+      "loss": 4.0606,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79536519163528e-05,
+      "loss": 4.0553,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794526596884228e-05,
+      "loss": 4.0401,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7936896400135486e-05,
+      "loss": 4.0394,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7928510452624966e-05,
+      "loss": 4.0466,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920124505114446e-05,
+      "loss": 4.0518,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911738557603926e-05,
+      "loss": 4.0478,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903352610093406e-05,
+      "loss": 4.054,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7894966662582886e-05,
+      "loss": 4.0513,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7886580715072366e-05,
+      "loss": 4.0424,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7878194767561846e-05,
+      "loss": 4.0422,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7869825198855055e-05,
+      "loss": 4.0376,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7861439251344535e-05,
+      "loss": 4.038,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785305330383402e-05,
+      "loss": 4.0479,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784468373512723e-05,
+      "loss": 4.0459,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783629778761671e-05,
+      "loss": 4.0413,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782791184010619e-05,
+      "loss": 4.0377,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781952589259567e-05,
+      "loss": 4.0315,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781115632388888e-05,
+      "loss": 4.0457,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780277037637836e-05,
+      "loss": 4.0369,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779438442886784e-05,
+      "loss": 4.0606,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778599848135732e-05,
+      "loss": 4.0489,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777762891265053e-05,
+      "loss": 4.0529,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776924296514001e-05,
+      "loss": 4.0425,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776085701762949e-05,
+      "loss": 4.0347,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7752471070118975e-05,
+      "loss": 4.0488,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744085122608455e-05,
+      "loss": 4.0497,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7735699175097935e-05,
+      "loss": 4.0411,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7727313227587415e-05,
+      "loss": 4.0347,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7718943658880624e-05,
+      "loss": 4.0483,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7710557711370104e-05,
+      "loss": 4.0369,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7702171763859584e-05,
+      "loss": 4.0401,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693785816349064e-05,
+      "loss": 4.0338,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768539986883854e-05,
+      "loss": 4.0455,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767703030013175e-05,
+      "loss": 4.0308,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766864435262123e-05,
+      "loss": 4.0482,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766025840511071e-05,
+      "loss": 4.0425,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765187245760019e-05,
+      "loss": 4.0585,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764348651008967e-05,
+      "loss": 4.0307,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763510056257915e-05,
+      "loss": 4.0533,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762671461506863e-05,
+      "loss": 4.0411,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761832866755811e-05,
+      "loss": 4.0505,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760995909885132e-05,
+      "loss": 4.0448,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76015731513408e-05,
+      "loss": 4.0384,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759318720383028e-05,
+      "loss": 4.0427,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758481763512349e-05,
+      "loss": 4.0583,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757644806641671e-05,
+      "loss": 4.0274,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756806211890618e-05,
+      "loss": 4.0505,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755967617139567e-05,
+      "loss": 4.0443,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755129022388515e-05,
+      "loss": 4.0398,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7542904276374627e-05,
+      "loss": 4.0371,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7534518328864107e-05,
+      "loss": 4.0392,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7526132381353586e-05,
+      "loss": 4.0464,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7517746433843066e-05,
+      "loss": 4.0385,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7509360486332546e-05,
+      "loss": 4.0387,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0687456130981445,
+      "eval_runtime": 316.6359,
+      "eval_samples_per_second": 1205.141,
+      "eval_steps_per_second": 37.662,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7500974538822026e-05,
+      "loss": 4.0377,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7492588591311506e-05,
+      "loss": 4.0334,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7484219022604715e-05,
+      "loss": 4.0339,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7475833075094195e-05,
+      "loss": 4.0437,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7467447127583675e-05,
+      "loss": 4.0516,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7459061180073155e-05,
+      "loss": 4.0371,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7450675232562635e-05,
+      "loss": 4.0358,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7442289285052115e-05,
+      "loss": 4.0339,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743391971634533e-05,
+      "loss": 4.0329,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742553376883481e-05,
+      "loss": 4.0479,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741714782132429e-05,
+      "loss": 4.0372,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740876187381377e-05,
+      "loss": 4.0418,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740039230510698e-05,
+      "loss": 4.0411,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739200635759646e-05,
+      "loss": 4.0314,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738362041008594e-05,
+      "loss": 4.0315,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737523446257542e-05,
+      "loss": 4.0301,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73668485150649e-05,
+      "loss": 4.0374,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735846256755437e-05,
+      "loss": 4.0344,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735009299884759e-05,
+      "loss": 4.0339,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734170705133707e-05,
+      "loss": 4.0397,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733332110382655e-05,
+      "loss": 4.0429,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732493515631603e-05,
+      "loss": 4.0455,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731654920880551e-05,
+      "loss": 4.0397,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730816326129499e-05,
+      "loss": 4.0326,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729977731378447e-05,
+      "loss": 4.0366,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729140774507768e-05,
+      "loss": 4.0427,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728302179756716e-05,
+      "loss": 4.0311,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727463585005664e-05,
+      "loss": 4.0351,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726624990254612e-05,
+      "loss": 4.0333,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72578639550356e-05,
+      "loss": 4.0256,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724947800752508e-05,
+      "loss": 4.0285,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724109206001456e-05,
+      "loss": 4.0384,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723270611250404e-05,
+      "loss": 4.0423,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722433654379725e-05,
+      "loss": 4.0362,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721595059628673e-05,
+      "loss": 4.0407,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720758102757994e-05,
+      "loss": 4.0306,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719919508006942e-05,
+      "loss": 4.0451,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71908091325589e-05,
+      "loss": 4.0296,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718242318504838e-05,
+      "loss": 4.0295,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717403723753786e-05,
+      "loss": 4.0237,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716565129002734e-05,
+      "loss": 4.0311,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715726534251682e-05,
+      "loss": 4.0301,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71488793950063e-05,
+      "loss": 4.0322,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714052620510324e-05,
+      "loss": 4.0317,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713214025759272e-05,
+      "loss": 4.0367,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712375431008221e-05,
+      "loss": 4.0283,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711536836257169e-05,
+      "loss": 4.0328,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710698241506117e-05,
+      "loss": 4.0268,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709859646755065e-05,
+      "loss": 4.0034,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709021052004013e-05,
+      "loss": 4.0424,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708182457252961e-05,
+      "loss": 4.0261,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707343862501909e-05,
+      "loss": 4.039,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7065069056312296e-05,
+      "loss": 4.0261,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7056683108801776e-05,
+      "loss": 4.0232,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7048313540094985e-05,
+      "loss": 4.0189,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7039927592584465e-05,
+      "loss": 4.0278,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7031541645073945e-05,
+      "loss": 4.0257,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7023155697563425e-05,
+      "loss": 4.0206,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7014769750052905e-05,
+      "loss": 4.0407,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7006383802542385e-05,
+      "loss": 4.0242,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6997997855031865e-05,
+      "loss": 4.0131,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6989611907521345e-05,
+      "loss": 4.0359,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698124233881456e-05,
+      "loss": 4.0236,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6972856391304034e-05,
+      "loss": 4.0203,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6964470443793514e-05,
+      "loss": 4.0326,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695611725389046e-05,
+      "loss": 4.0249,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694773130637994e-05,
+      "loss": 4.0199,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693934535886942e-05,
+      "loss": 4.0322,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69309594113589e-05,
+      "loss": 4.0151,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692257346384838e-05,
+      "loss": 4.0203,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691418751633786e-05,
+      "loss": 4.0329,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690580156882734e-05,
+      "loss": 4.0305,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689741562131682e-05,
+      "loss": 4.0334,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68890296738063e-05,
+      "loss": 4.0263,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688064372629578e-05,
+      "loss": 4.0516,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687227415758899e-05,
+      "loss": 4.0203,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686388821007847e-05,
+      "loss": 4.0337,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685550226256795e-05,
+      "loss": 4.0266,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684711631505743e-05,
+      "loss": 4.027,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683873036754691e-05,
+      "loss": 4.023,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6830360798840116e-05,
+      "loss": 4.039,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6821974851329596e-05,
+      "loss": 4.03,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681358890381908e-05,
+      "loss": 4.0436,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680520295630856e-05,
+      "loss": 4.0119,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679681700879804e-05,
+      "loss": 4.0186,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678843106128752e-05,
+      "loss": 4.0265,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678006149258073e-05,
+      "loss": 4.0354,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677167554507021e-05,
+      "loss": 4.0246,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676330597636342e-05,
+      "loss": 4.0314,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.67549200288529e-05,
+      "loss": 4.0225,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674653408134238e-05,
+      "loss": 4.0206,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673814813383186e-05,
+      "loss": 4.0232,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672976218632134e-05,
+      "loss": 4.021,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672137623881082e-05,
+      "loss": 4.0317,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67129902913003e-05,
+      "loss": 4.0421,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670460434378978e-05,
+      "loss": 4.037,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669621839627926e-05,
+      "loss": 4.0213,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668783244876875e-05,
+      "loss": 4.018,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6679462880061956e-05,
+      "loss": 4.0251,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6671076932551436e-05,
+      "loss": 4.0279,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6662690985040916e-05,
+      "loss": 4.0303,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6654305037530396e-05,
+      "loss": 4.028,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664591909001987e-05,
+      "loss": 4.0313,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663753314250935e-05,
+      "loss": 4.0203,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662914719499883e-05,
+      "loss": 4.021,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662077762629204e-05,
+      "loss": 4.0175,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661239167878152e-05,
+      "loss": 4.0185,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6604005731271e-05,
+      "loss": 4.0273,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6595619783760485e-05,
+      "loss": 4.0277,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6587233836249965e-05,
+      "loss": 4.0163,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6578847888739445e-05,
+      "loss": 4.0184,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6570478320032654e-05,
+      "loss": 4.0081,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6562092372522134e-05,
+      "loss": 4.0284,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6553706425011614e-05,
+      "loss": 4.0192,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6545320477501094e-05,
+      "loss": 4.0335,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6536934529990574e-05,
+      "loss": 4.0302,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6528548582480054e-05,
+      "loss": 4.0316,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6520162634969534e-05,
+      "loss": 4.022,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6511776687459014e-05,
+      "loss": 4.0182,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650340711875222e-05,
+      "loss": 4.0248,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64950211712417e-05,
+      "loss": 4.0304,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648663522373118e-05,
+      "loss": 4.0198,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647824927622067e-05,
+      "loss": 4.0126,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646987970751388e-05,
+      "loss": 4.028,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646149376000336e-05,
+      "loss": 4.0182,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645310781249284e-05,
+      "loss": 4.019,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644473824378605e-05,
+      "loss": 4.0122,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643635229627553e-05,
+      "loss": 4.029,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642796634876501e-05,
+      "loss": 4.0083,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641958040125449e-05,
+      "loss": 4.0298,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6411210832547697e-05,
+      "loss": 4.0236,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6402824885037176e-05,
+      "loss": 4.0347,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6394438937526656e-05,
+      "loss": 4.0099,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6386052990016136e-05,
+      "loss": 4.0312,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637766704250562e-05,
+      "loss": 4.0269,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63692810949951e-05,
+      "loss": 4.0255,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636089514748458e-05,
+      "loss": 4.0261,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6352509199974056e-05,
+      "loss": 4.018,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6344123252463536e-05,
+      "loss": 4.0208,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6335737304953016e-05,
+      "loss": 4.0379,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6327351357442496e-05,
+      "loss": 4.013,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6318981788735705e-05,
+      "loss": 4.0263,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6310595841225185e-05,
+      "loss": 4.0275,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6302209893714665e-05,
+      "loss": 4.0178,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6293823946204145e-05,
+      "loss": 4.0188,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6285437998693625e-05,
+      "loss": 4.0213,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6277052051183105e-05,
+      "loss": 4.0247,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626868248247632e-05,
+      "loss": 4.022,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.62602965349658e-05,
+      "loss": 4.0176,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.056021213531494,
+      "eval_runtime": 339.4471,
+      "eval_samples_per_second": 1124.155,
+      "eval_steps_per_second": 35.131,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625191058745528e-05,
+      "loss": 4.0136,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624352463994476e-05,
+      "loss": 4.0128,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623513869243424e-05,
+      "loss": 4.0135,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622675274492372e-05,
+      "loss": 4.025,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62183667974132e-05,
+      "loss": 4.0281,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620998084990268e-05,
+      "loss": 4.0201,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620159490239216e-05,
+      "loss": 4.0172,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619320895488164e-05,
+      "loss": 4.017,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6184823007371114e-05,
+      "loss": 4.0096,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6176437059860594e-05,
+      "loss": 4.0251,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6168051112350074e-05,
+      "loss": 4.0203,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6159665164839553e-05,
+      "loss": 4.0255,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615127921732904e-05,
+      "loss": 4.0209,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614289326981852e-05,
+      "loss": 4.0108,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6134507322308e-05,
+      "loss": 4.0147,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612612137479748e-05,
+      "loss": 4.0046,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611773542728696e-05,
+      "loss": 4.0208,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610934947977644e-05,
+      "loss": 4.0168,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610097991106965e-05,
+      "loss": 4.0134,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609259396355913e-05,
+      "loss": 4.0183,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608420801604861e-05,
+      "loss": 4.0247,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607582206853809e-05,
+      "loss": 4.0257,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606743612102757e-05,
+      "loss": 4.0214,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605906655232078e-05,
+      "loss": 4.0178,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605068060481026e-05,
+      "loss": 4.0173,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604229465729974e-05,
+      "loss": 4.0208,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6033908709789225e-05,
+      "loss": 4.0086,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6025522762278705e-05,
+      "loss": 4.0152,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6017153193571914e-05,
+      "loss": 4.013,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6008767246061394e-05,
+      "loss": 4.008,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6000381298550874e-05,
+      "loss": 4.0125,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5991995351040354e-05,
+      "loss": 4.0189,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5983609403529834e-05,
+      "loss": 4.015,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597522345601931e-05,
+      "loss": 4.0277,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596683750850879e-05,
+      "loss": 4.0138,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595845156099827e-05,
+      "loss": 4.0142,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5950081992291476e-05,
+      "loss": 4.0265,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594171242358469e-05,
+      "loss": 4.0056,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593332647607418e-05,
+      "loss": 4.0109,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592494052856366e-05,
+      "loss": 4.0059,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591655458105313e-05,
+      "loss": 4.0124,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590816863354261e-05,
+      "loss": 4.0112,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589978268603209e-05,
+      "loss": 4.0128,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589139673852157e-05,
+      "loss": 4.0148,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588302716981478e-05,
+      "loss": 4.0151,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587464122230426e-05,
+      "loss": 4.0151,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5866271653597476e-05,
+      "loss": 4.0086,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585788570608695e-05,
+      "loss": 4.0091,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584949975857643e-05,
+      "loss": 3.9845,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5841113811065916e-05,
+      "loss": 4.023,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583274424235913e-05,
+      "loss": 4.0072,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5824358294848605e-05,
+      "loss": 4.0186,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5815972347338085e-05,
+      "loss": 4.0083,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5807586399827565e-05,
+      "loss": 4.0028,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5799200452317045e-05,
+      "loss": 4.0011,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5790830883610254e-05,
+      "loss": 4.0079,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5782444936099734e-05,
+      "loss": 4.0106,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5774058988589214e-05,
+      "loss": 3.9987,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5765673041078694e-05,
+      "loss": 4.0225,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5757287093568174e-05,
+      "loss": 4.0081,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5748901146057654e-05,
+      "loss": 3.9909,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5740515198547134e-05,
+      "loss": 4.0172,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5732129251036614e-05,
+      "loss": 4.0062,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572375968232983e-05,
+      "loss": 4.0022,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571539011362304e-05,
+      "loss": 4.0117,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570700416611252e-05,
+      "loss": 4.0087,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5698618218602e-05,
+      "loss": 4.0001,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569023227109148e-05,
+      "loss": 4.0156,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568186270238469e-05,
+      "loss": 3.9966,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567347675487417e-05,
+      "loss": 3.9987,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566509080736365e-05,
+      "loss": 4.017,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565670485985313e-05,
+      "loss": 4.0098,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564831891234261e-05,
+      "loss": 4.0147,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563993296483209e-05,
+      "loss": 4.0118,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563154701732157e-05,
+      "loss": 4.0274,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562316106981105e-05,
+      "loss": 4.0058,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5614775122300534e-05,
+      "loss": 4.0181,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5606389174790014e-05,
+      "loss": 4.0073,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5598003227279494e-05,
+      "loss": 4.0099,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55896336585727e-05,
+      "loss": 4.0049,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558126408986591e-05,
+      "loss": 4.0189,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557287814235539e-05,
+      "loss": 4.0142,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556449219484487e-05,
+      "loss": 4.0281,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555610624733435e-05,
+      "loss": 3.9953,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554772029982383e-05,
+      "loss": 3.9983,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553935073111704e-05,
+      "loss": 4.0091,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553096478360652e-05,
+      "loss": 4.0157,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552259521489974e-05,
+      "loss": 4.005,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5514225646192946e-05,
+      "loss": 4.0164,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5505839698682426e-05,
+      "loss": 4.0043,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5497453751171906e-05,
+      "loss": 3.9991,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5489067803661386e-05,
+      "loss": 4.006,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5480681856150866e-05,
+      "loss": 4.0048,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5472295908640346e-05,
+      "loss": 4.0152,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5463909961129826e-05,
+      "loss": 4.0183,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5455524013619306e-05,
+      "loss": 4.021,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5447138066108786e-05,
+      "loss": 3.9982,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543875211859826e-05,
+      "loss": 4.0034,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543036617108774e-05,
+      "loss": 4.0071,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5421980223577226e-05,
+      "loss": 4.0119,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5413594276066706e-05,
+      "loss": 4.0102,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5405208328556186e-05,
+      "loss": 4.0128,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5396838759849395e-05,
+      "loss": 4.0122,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5388452812338875e-05,
+      "loss": 4.0043,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5380066864828355e-05,
+      "loss": 4.0062,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5371680917317835e-05,
+      "loss": 3.9968,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5363294969807314e-05,
+      "loss": 4.0037,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5354925401100524e-05,
+      "loss": 4.0072,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5346539453590004e-05,
+      "loss": 4.0093,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5338153506079483e-05,
+      "loss": 3.9997,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5329767558568963e-05,
+      "loss": 4.0035,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532138161105844e-05,
+      "loss": 3.9884,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531299566354792e-05,
+      "loss": 4.0151,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530460971603741e-05,
+      "loss": 4.0019,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529624014733062e-05,
+      "loss": 4.0179,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52878541998201e-05,
+      "loss": 4.0102,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527946825230958e-05,
+      "loss": 4.0138,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527108230479906e-05,
+      "loss": 4.0029,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526269635728854e-05,
+      "loss": 3.999,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525431040977802e-05,
+      "loss": 4.012,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524594084107123e-05,
+      "loss": 4.0099,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523755489356071e-05,
+      "loss": 4.0057,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522916894605019e-05,
+      "loss": 3.9985,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522078299853967e-05,
+      "loss": 4.0072,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521239705102915e-05,
+      "loss": 4.0019,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5204027482322364e-05,
+      "loss": 4.0033,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5195641534811844e-05,
+      "loss": 3.9939,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5187255587301324e-05,
+      "loss": 4.0082,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51788696397908e-05,
+      "loss": 3.9911,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517048369228028e-05,
+      "loss": 4.015,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516211412357349e-05,
+      "loss": 4.0056,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515372817606297e-05,
+      "loss": 4.0171,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5145342228552446e-05,
+      "loss": 3.9964,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5136956281041926e-05,
+      "loss": 4.0154,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5128570333531406e-05,
+      "loss": 4.0057,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5120184386020886e-05,
+      "loss": 4.0086,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5111798438510366e-05,
+      "loss": 4.0099,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5103412490999846e-05,
+      "loss": 3.9992,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5095026543489326e-05,
+      "loss": 4.0057,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508665697478254e-05,
+      "loss": 4.0194,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507827102727202e-05,
+      "loss": 3.9977,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50698850797615e-05,
+      "loss": 4.0098,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506149913225098e-05,
+      "loss": 4.009,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505312956354419e-05,
+      "loss": 4.0019,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504474361603367e-05,
+      "loss": 4.0031,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503635766852315e-05,
+      "loss": 4.0028,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502797172101263e-05,
+      "loss": 4.0049,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501958577350211e-05,
+      "loss": 4.0051,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501121620479532e-05,
+      "loss": 3.9991,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.045508861541748,
+      "eval_runtime": 307.4695,
+      "eval_samples_per_second": 1241.069,
+      "eval_steps_per_second": 38.784,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.50028302572848e-05,
+      "loss": 4.0051,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499444430977428e-05,
+      "loss": 3.9939,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4986058362263766e-05,
+      "loss": 3.9968,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4977688793556975e-05,
+      "loss": 4.0118,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4969302846046455e-05,
+      "loss": 4.0062,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4960916898535935e-05,
+      "loss": 4.0048,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4952530951025415e-05,
+      "loss": 4.0041,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4944161382318624e-05,
+      "loss": 3.9979,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4935775434808104e-05,
+      "loss": 3.9947,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4927389487297584e-05,
+      "loss": 4.0027,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4919003539787064e-05,
+      "loss": 4.0076,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491063397108027e-05,
+      "loss": 4.0051,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490226440237349e-05,
+      "loss": 4.0063,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48938948336667e-05,
+      "loss": 3.9913,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488550888615618e-05,
+      "loss": 4.0,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487712293864566e-05,
+      "loss": 3.9882,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486875336993887e-05,
+      "loss": 4.0038,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486036742242835e-05,
+      "loss": 4.0005,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.485198147491783e-05,
+      "loss": 3.9942,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.484359552740731e-05,
+      "loss": 4.004,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.483520957989679e-05,
+      "loss": 4.0072,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482682363238627e-05,
+      "loss": 4.0098,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481843768487575e-05,
+      "loss": 4.0037,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4810051737365227e-05,
+      "loss": 3.9991,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4801665789854707e-05,
+      "loss": 4.0046,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4793279842344187e-05,
+      "loss": 4.0037,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478489389483367e-05,
+      "loss": 3.9947,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.477650794732315e-05,
+      "loss": 3.9981,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476813837861636e-05,
+      "loss": 3.996,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.475975243110584e-05,
+      "loss": 3.9906,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475136648359532e-05,
+      "loss": 3.9981,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.47429805360848e-05,
+      "loss": 3.9983,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.473459458857428e-05,
+      "loss": 3.9993,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4726208641063755e-05,
+      "loss": 4.0101,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4717822693553235e-05,
+      "loss": 3.9993,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470945312484645e-05,
+      "loss": 3.9989,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470106717733593e-05,
+      "loss": 4.0074,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469268122982541e-05,
+      "loss": 3.9884,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468429528231489e-05,
+      "loss": 3.9938,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467592571360811e-05,
+      "loss": 3.9959,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466753976609758e-05,
+      "loss": 3.9911,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465915381858706e-05,
+      "loss": 3.9929,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465076787107654e-05,
+      "loss": 3.9965,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464238192356602e-05,
+      "loss": 3.9988,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46339959760555e-05,
+      "loss": 3.9958,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462561002854498e-05,
+      "loss": 3.9998,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461722408103446e-05,
+      "loss": 3.9948,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460885451232767e-05,
+      "loss": 3.9953,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460046856481715e-05,
+      "loss": 3.9708,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4592098996110365e-05,
+      "loss": 4.0068,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4583713048599845e-05,
+      "loss": 3.99,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4575327101089325e-05,
+      "loss": 3.9991,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4566941153578805e-05,
+      "loss": 3.9923,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4558555206068285e-05,
+      "loss": 3.9872,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4550185637361494e-05,
+      "loss": 3.9858,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45418160686547e-05,
+      "loss": 3.99,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.453343012114418e-05,
+      "loss": 3.9969,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452504417363366e-05,
+      "loss": 3.9822,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451665822612314e-05,
+      "loss": 4.0075,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450827227861262e-05,
+      "loss": 3.9904,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44998863311021e-05,
+      "loss": 3.9769,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449150038359158e-05,
+      "loss": 3.9989,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448311443608106e-05,
+      "loss": 3.9862,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447472848857054e-05,
+      "loss": 3.9896,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 3.9913,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 3.995,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444958702484272e-05,
+      "loss": 3.988,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44412010773322e-05,
+      "loss": 3.9988,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443283150862541e-05,
+      "loss": 3.9785,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442444556111489e-05,
+      "loss": 3.9844,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441605961360437e-05,
+      "loss": 3.9968,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440767366609385e-05,
+      "loss": 3.9958,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439928771858333e-05,
+      "loss": 3.999,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439090177107281e-05,
+      "loss": 3.9919,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438251582356229e-05,
+      "loss": 4.0154,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.437412987605177e-05,
+      "loss": 3.9883,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436576030734498e-05,
+      "loss": 4.0028,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435737435983446e-05,
+      "loss": 3.988,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434898841232394e-05,
+      "loss": 3.9941,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4340602464813416e-05,
+      "loss": 3.9852,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4332216517302896e-05,
+      "loss": 4.0014,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4323830569792376e-05,
+      "loss": 3.9994,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4315444622281856e-05,
+      "loss": 4.0116,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4307058674771336e-05,
+      "loss": 3.9835,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4298689106064545e-05,
+      "loss": 3.9845,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4290303158554025e-05,
+      "loss": 3.9894,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4281917211043505e-05,
+      "loss": 4.0024,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427354764233672e-05,
+      "loss": 3.989,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.42651616948262e-05,
+      "loss": 4.0013,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425677574731568e-05,
+      "loss": 3.9896,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424838979980516e-05,
+      "loss": 3.9841,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424000385229464e-05,
+      "loss": 3.9913,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423163428358785e-05,
+      "loss": 3.9838,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422324833607733e-05,
+      "loss": 4.0003,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421486238856681e-05,
+      "loss": 4.0055,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420649281986002e-05,
+      "loss": 4.0039,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41981068723495e-05,
+      "loss": 3.9838,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418972092483898e-05,
+      "loss": 3.9858,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418133497732846e-05,
+      "loss": 3.9917,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.417294902981794e-05,
+      "loss": 3.9939,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416456308230742e-05,
+      "loss": 3.9975,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4156177134796905e-05,
+      "loss": 3.9937,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4147791187286385e-05,
+      "loss": 3.9946,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4139421618579594e-05,
+      "loss": 3.9866,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4131035671069074e-05,
+      "loss": 3.9936,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4122649723558554e-05,
+      "loss": 3.9783,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4114263776048034e-05,
+      "loss": 3.9872,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4105877828537514e-05,
+      "loss": 3.9895,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4097491881026994e-05,
+      "loss": 3.9935,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4089105933516474e-05,
+      "loss": 3.9856,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408071998600595e-05,
+      "loss": 3.9895,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407233403849543e-05,
+      "loss": 3.9686,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406396446978864e-05,
+      "loss": 3.9942,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405557852227812e-05,
+      "loss": 3.9902,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40471925747676e-05,
+      "loss": 3.9945,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403880662725708e-05,
+      "loss": 3.997,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403045343735403e-05,
+      "loss": 3.9986,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402206748984351e-05,
+      "loss": 3.9879,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401368154233299e-05,
+      "loss": 3.9849,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.400529559482247e-05,
+      "loss": 3.9963,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.399692602611568e-05,
+      "loss": 3.9928,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3988540078605157e-05,
+      "loss": 3.9878,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3980154131094637e-05,
+      "loss": 3.9858,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3971768183584117e-05,
+      "loss": 3.9892,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3963382236073596e-05,
+      "loss": 3.9889,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3954996288563076e-05,
+      "loss": 3.9839,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3946610341052556e-05,
+      "loss": 3.9833,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3938224393542036e-05,
+      "loss": 3.9898,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392985482483525e-05,
+      "loss": 3.9768,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3921468877324725e-05,
+      "loss": 3.9973,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3913082929814205e-05,
+      "loss": 3.9888,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3904696982303685e-05,
+      "loss": 4.0024,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3896311034793165e-05,
+      "loss": 3.9775,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3887941466086374e-05,
+      "loss": 4.0015,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3879555518575854e-05,
+      "loss": 3.9944,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3871169571065334e-05,
+      "loss": 3.9908,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3862783623554814e-05,
+      "loss": 3.9948,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3854397676044294e-05,
+      "loss": 3.9838,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3846011728533774e-05,
+      "loss": 3.9954,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383764215982699e-05,
+      "loss": 3.9998,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382925621231647e-05,
+      "loss": 3.9827,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382087026480595e-05,
+      "loss": 3.9921,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381248431729543e-05,
+      "loss": 3.9935,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.380411474858864e-05,
+      "loss": 3.9845,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379572880107812e-05,
+      "loss": 3.9867,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.37873428535676e-05,
+      "loss": 3.9911,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377895690605708e-05,
+      "loss": 3.9865,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 3.9923,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376220138983977e-05,
+      "loss": 3.9839,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.036524295806885,
+      "eval_runtime": 315.6959,
+      "eval_samples_per_second": 1208.73,
+      "eval_steps_per_second": 37.774,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375381544232925e-05,
+      "loss": 4.0083,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374542949481873e-05,
+      "loss": 3.9777,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3737043547308215e-05,
+      "loss": 3.9791,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3728657599797694e-05,
+      "loss": 3.9975,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3720271652287174e-05,
+      "loss": 3.99,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3711885704776654e-05,
+      "loss": 3.9897,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3703499757266134e-05,
+      "loss": 3.9863,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3695113809755614e-05,
+      "loss": 3.9847,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368672786224509e-05,
+      "loss": 3.9758,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367834191473457e-05,
+      "loss": 3.9926,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366995596722405e-05,
+      "loss": 3.9897,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366157001971353e-05,
+      "loss": 3.9878,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365318407220301e-05,
+      "loss": 3.9941,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364479812469249e-05,
+      "loss": 3.9764,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.363641217718197e-05,
+      "loss": 3.9814,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362802622967145e-05,
+      "loss": 3.9759,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361964028216093e-05,
+      "loss": 3.9882,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361127071345414e-05,
+      "loss": 3.9845,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360288476594362e-05,
+      "loss": 3.9804,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35944988184331e-05,
+      "loss": 3.982,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358611287092258e-05,
+      "loss": 3.9946,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357772692341206e-05,
+      "loss": 3.9911,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356935735470527e-05,
+      "loss": 3.9894,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356097140719475e-05,
+      "loss": 3.9855,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355258545968423e-05,
+      "loss": 3.9866,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354419951217371e-05,
+      "loss": 3.9855,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353582994346692e-05,
+      "loss": 3.9822,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35274439959564e-05,
+      "loss": 3.983,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351905804844588e-05,
+      "loss": 3.979,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351067210093537e-05,
+      "loss": 3.9761,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350228615342485e-05,
+      "loss": 3.982,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349390020591433e-05,
+      "loss": 3.9844,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.34855142584038e-05,
+      "loss": 3.9889,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347712831089328e-05,
+      "loss": 3.9906,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346874236338276e-05,
+      "loss": 3.9841,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346035641587224e-05,
+      "loss": 3.9881,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345197046836172e-05,
+      "loss": 3.9942,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344360089965493e-05,
+      "loss": 3.9698,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343521495214441e-05,
+      "loss": 3.9788,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342682900463389e-05,
+      "loss": 3.9799,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341844305712337e-05,
+      "loss": 3.9764,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3410073488416585e-05,
+      "loss": 3.9788,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3401687540906065e-05,
+      "loss": 3.9826,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3393301593395545e-05,
+      "loss": 3.9863,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3384915645885025e-05,
+      "loss": 3.9784,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3376529698374505e-05,
+      "loss": 3.9822,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3368143750863985e-05,
+      "loss": 3.9805,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359757803353465e-05,
+      "loss": 3.9771,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351388234646674e-05,
+      "loss": 3.9576,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3343002287136154e-05,
+      "loss": 3.9885,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3334616339625634e-05,
+      "loss": 3.9771,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332624677091884e-05,
+      "loss": 3.9857,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331786082340832e-05,
+      "loss": 3.9769,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33094748758978e-05,
+      "loss": 3.9678,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330108892838728e-05,
+      "loss": 3.9717,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32927193596805e-05,
+      "loss": 3.9775,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328433341216998e-05,
+      "loss": 3.9835,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327594746465946e-05,
+      "loss": 3.964,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326756151714894e-05,
+      "loss": 3.994,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325917556963842e-05,
+      "loss": 3.9772,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32507896221279e-05,
+      "loss": 3.9621,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324240367461738e-05,
+      "loss": 3.9849,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323401772710686e-05,
+      "loss": 3.9699,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322564815840007e-05,
+      "loss": 3.9773,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321726221088955e-05,
+      "loss": 3.9698,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320889264218276e-05,
+      "loss": 3.9842,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320050669467224e-05,
+      "loss": 3.9691,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319212074716172e-05,
+      "loss": 3.9847,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31837347996512e-05,
+      "loss": 3.9629,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317536523094441e-05,
+      "loss": 3.972,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316697928343389e-05,
+      "loss": 3.9803,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315859333592337e-05,
+      "loss": 3.9863,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315020738841285e-05,
+      "loss": 3.9801,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314182144090233e-05,
+      "loss": 3.978,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133435493391805e-05,
+      "loss": 3.9998,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3125049545881285e-05,
+      "loss": 3.9743,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31166799771745e-05,
+      "loss": 3.9842,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310829402966398e-05,
+      "loss": 3.9772,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309990808215346e-05,
+      "loss": 3.9806,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309152213464294e-05,
+      "loss": 3.9755,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308313618713242e-05,
+      "loss": 3.9821,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30747502396219e-05,
+      "loss": 3.9819,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306636429211138e-05,
+      "loss": 3.9938,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305797834460086e-05,
+      "loss": 3.9764,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304959239709034e-05,
+      "loss": 3.9669,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304123920718728e-05,
+      "loss": 3.9736,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303285325967676e-05,
+      "loss": 3.983,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302446731216624e-05,
+      "loss": 3.978,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301608136465572e-05,
+      "loss": 3.9858,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30076954171452e-05,
+      "loss": 3.9746,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299930946963468e-05,
+      "loss": 3.9678,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299092352212416e-05,
+      "loss": 3.9752,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2982537574613646e-05,
+      "loss": 3.9722,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2974168005906855e-05,
+      "loss": 3.9885,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2965782058396335e-05,
+      "loss": 3.9823,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2957396110885815e-05,
+      "loss": 3.9918,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2949010163375295e-05,
+      "loss": 3.9713,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2940640594668504e-05,
+      "loss": 3.975,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293227102596171e-05,
+      "loss": 3.9766,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292388507845119e-05,
+      "loss": 3.9787,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291549913094067e-05,
+      "loss": 3.9873,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290711318343015e-05,
+      "loss": 3.9774,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289872723591963e-05,
+      "loss": 3.9824,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289034128840911e-05,
+      "loss": 3.97,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288195534089859e-05,
+      "loss": 3.9825,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287356939338808e-05,
+      "loss": 3.9615,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286519982468129e-05,
+      "loss": 3.9732,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285681387717077e-05,
+      "loss": 3.975,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284842792966025e-05,
+      "loss": 3.981,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284004198214973e-05,
+      "loss": 3.971,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283167241344294e-05,
+      "loss": 3.9767,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282328646593242e-05,
+      "loss": 3.9555,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28149005184219e-05,
+      "loss": 3.9844,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280651457091138e-05,
+      "loss": 3.9712,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2798145002204586e-05,
+      "loss": 3.9804,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2789759054694066e-05,
+      "loss": 3.9848,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2781373107183546e-05,
+      "loss": 3.9846,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277298715967303e-05,
+      "loss": 3.9717,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276461759096624e-05,
+      "loss": 3.9717,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275623164345572e-05,
+      "loss": 3.9814,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27478456959452e-05,
+      "loss": 3.9747,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273945974843468e-05,
+      "loss": 3.9774,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273107380092416e-05,
+      "loss": 3.97,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.272270423221737e-05,
+      "loss": 3.9765,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271431828470685e-05,
+      "loss": 3.9698,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270594871600006e-05,
+      "loss": 3.9666,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269756276848954e-05,
+      "loss": 3.9719,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268917682097902e-05,
+      "loss": 3.9734,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26807908734685e-05,
+      "loss": 3.9649,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2672404925957987e-05,
+      "loss": 3.9841,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2664018978447467e-05,
+      "loss": 3.9758,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265563303093694e-05,
+      "loss": 3.9861,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264724708342642e-05,
+      "loss": 3.969,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2638877514719636e-05,
+      "loss": 3.983,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2630491567209115e-05,
+      "loss": 3.9808,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262210561969859e-05,
+      "loss": 3.9771,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261371967218807e-05,
+      "loss": 3.9754,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2605366482285014e-05,
+      "loss": 3.971,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2596980534774494e-05,
+      "loss": 3.9812,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2588594587263974e-05,
+      "loss": 3.9858,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2580208639753453e-05,
+      "loss": 3.9721,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257182269224294e-05,
+      "loss": 3.9748,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2563436744732413e-05,
+      "loss": 3.9812,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255505079722189e-05,
+      "loss": 3.9726,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254666484971137e-05,
+      "loss": 3.972,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253827890220085e-05,
+      "loss": 3.9784,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252990933349406e-05,
+      "loss": 3.9726,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252152338598354e-05,
+      "loss": 3.9792,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251315381727676e-05,
+      "loss": 3.9667,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.027721881866455,
+      "eval_runtime": 336.7649,
+      "eval_samples_per_second": 1133.108,
+      "eval_steps_per_second": 35.41,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250476786976624e-05,
+      "loss": 3.9751,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249638192225571e-05,
+      "loss": 3.9657,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248799597474519e-05,
+      "loss": 3.9653,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247961002723468e-05,
+      "loss": 3.9795,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247122407972416e-05,
+      "loss": 3.9801,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246283813221364e-05,
+      "loss": 3.9744,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245445218470312e-05,
+      "loss": 3.9699,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244608261599633e-05,
+      "loss": 3.9702,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243769666848581e-05,
+      "loss": 3.963,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2429327099779016e-05,
+      "loss": 3.9808,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2420941152268496e-05,
+      "loss": 3.9742,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2412555204757976e-05,
+      "loss": 3.9783,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2404169257247456e-05,
+      "loss": 3.9744,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2395783309736936e-05,
+      "loss": 3.9644,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2387397362226416e-05,
+      "loss": 3.9703,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2379011414715896e-05,
+      "loss": 3.9613,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237064184600911e-05,
+      "loss": 3.975,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236225589849859e-05,
+      "loss": 3.9679,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235386995098807e-05,
+      "loss": 3.9695,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234548400347755e-05,
+      "loss": 3.9685,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233711443477076e-05,
+      "loss": 3.9787,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232872848726024e-05,
+      "loss": 3.9765,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232034253974972e-05,
+      "loss": 3.9741,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23119565922392e-05,
+      "loss": 3.9708,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.230357064472868e-05,
+      "loss": 3.978,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.229518469721816e-05,
+      "loss": 3.9683,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228679874970764e-05,
+      "loss": 3.9651,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227841280219712e-05,
+      "loss": 3.9743,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227004323349033e-05,
+      "loss": 3.9625,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2261657285979816e-05,
+      "loss": 3.9618,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2253271338469296e-05,
+      "loss": 3.9665,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2244885390958776e-05,
+      "loss": 3.9655,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2236515822251985e-05,
+      "loss": 3.9775,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2228129874741465e-05,
+      "loss": 3.9802,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2219743927230945e-05,
+      "loss": 3.9652,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2211357979720425e-05,
+      "loss": 3.9791,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.22029720322099e-05,
+      "loss": 3.9768,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2194602463503114e-05,
+      "loss": 3.9581,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2186216515992594e-05,
+      "loss": 3.9661,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217783056848207e-05,
+      "loss": 3.965,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2169444620971554e-05,
+      "loss": 3.9603,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2161058673461034e-05,
+      "loss": 3.9673,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215268910475425e-05,
+      "loss": 3.9747,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214430315724372e-05,
+      "loss": 3.9657,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21359172097332e-05,
+      "loss": 3.9657,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212753126222268e-05,
+      "loss": 3.9708,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211914531471216e-05,
+      "loss": 3.9672,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211075936720164e-05,
+      "loss": 3.9653,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210238979849485e-05,
+      "loss": 3.9445,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209400385098433e-05,
+      "loss": 3.9736,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208561790347381e-05,
+      "loss": 3.9677,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207723195596329e-05,
+      "loss": 3.9742,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20688623872565e-05,
+      "loss": 3.9611,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206047643974599e-05,
+      "loss": 3.9561,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205209049223547e-05,
+      "loss": 3.9598,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204370454472495e-05,
+      "loss": 3.9633,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203531859721443e-05,
+      "loss": 3.9681,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202693264970391e-05,
+      "loss": 3.9479,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201854670219339e-05,
+      "loss": 3.9774,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201016075468287e-05,
+      "loss": 3.9651,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.200177480717235e-05,
+      "loss": 3.9529,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199338885966183e-05,
+      "loss": 3.9697,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1985019290955036e-05,
+      "loss": 3.9585,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1976633343444516e-05,
+      "loss": 3.9619,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1968247395933996e-05,
+      "loss": 3.9563,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1959861448423476e-05,
+      "loss": 3.9716,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1951491879716685e-05,
+      "loss": 3.9592,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19431223110099e-05,
+      "loss": 3.9644,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.193473636349938e-05,
+      "loss": 3.9561,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192635041598886e-05,
+      "loss": 3.958,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191796446847834e-05,
+      "loss": 3.9632,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190957852096782e-05,
+      "loss": 3.9787,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19011925734573e-05,
+      "loss": 3.9646,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189280662594678e-05,
+      "loss": 3.9639,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188442067843626e-05,
+      "loss": 3.9839,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18760674885332e-05,
+      "loss": 3.9618,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186768154102268e-05,
+      "loss": 3.9692,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185929559351216e-05,
+      "loss": 3.9673,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185090964600164e-05,
+      "loss": 3.9716,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1842523698491126e-05,
+      "loss": 3.96,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1834137750980606e-05,
+      "loss": 3.9666,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1825751803470086e-05,
+      "loss": 3.9717,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181736585595956e-05,
+      "loss": 3.982,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1808996287252775e-05,
+      "loss": 3.9645,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1800610339742255e-05,
+      "loss": 3.9512,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1792224392231735e-05,
+      "loss": 3.9608,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.178383844472121e-05,
+      "loss": 3.9716,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177545249721069e-05,
+      "loss": 3.9659,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176706654970017e-05,
+      "loss": 3.9691,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.175869698099338e-05,
+      "loss": 3.9629,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1750311033482863e-05,
+      "loss": 3.9555,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174192508597234e-05,
+      "loss": 3.961,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173353913846182e-05,
+      "loss": 3.9586,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17251531909513e-05,
+      "loss": 3.9722,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171676724344078e-05,
+      "loss": 3.9684,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170838129593026e-05,
+      "loss": 3.9821,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169999534841974e-05,
+      "loss": 3.9549,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169162577971295e-05,
+      "loss": 3.9637,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168323983220243e-05,
+      "loss": 3.9647,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167485388469191e-05,
+      "loss": 3.9671,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166648431598512e-05,
+      "loss": 3.9713,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16580983684746e-05,
+      "loss": 3.9655,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164971242096408e-05,
+      "loss": 3.9692,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16413428522573e-05,
+      "loss": 3.9549,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163295690474678e-05,
+      "loss": 3.9709,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162457095723626e-05,
+      "loss": 3.9474,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161618500972574e-05,
+      "loss": 3.9606,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160779906221522e-05,
+      "loss": 3.9637,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15994131147047e-05,
+      "loss": 3.9699,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159102716719418e-05,
+      "loss": 3.962,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158264121968366e-05,
+      "loss": 3.9594,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1574271650976866e-05,
+      "loss": 3.9527,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1565885703466346e-05,
+      "loss": 3.9643,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1557516134759555e-05,
+      "loss": 3.9599,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1549130187249035e-05,
+      "loss": 3.9696,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1540744239738515e-05,
+      "loss": 3.9731,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1532358292227995e-05,
+      "loss": 3.9718,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152397234471748e-05,
+      "loss": 3.9574,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151558639720696e-05,
+      "loss": 3.9608,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150720044969644e-05,
+      "loss": 3.9706,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149883088098965e-05,
+      "loss": 3.9636,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149044493347913e-05,
+      "loss": 3.964,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148205898596861e-05,
+      "loss": 3.961,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147367303845809e-05,
+      "loss": 3.9622,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1465287090947564e-05,
+      "loss": 3.9601,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145691752224078e-05,
+      "loss": 3.9541,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144853157473026e-05,
+      "loss": 3.963,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144014562721974e-05,
+      "loss": 3.9603,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143175967970922e-05,
+      "loss": 3.9549,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14233737321987e-05,
+      "loss": 3.9645,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.141498778468818e-05,
+      "loss": 3.967,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140660183717766e-05,
+      "loss": 3.9709,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139823226847087e-05,
+      "loss": 3.9636,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138984632096035e-05,
+      "loss": 3.9677,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138146037344983e-05,
+      "loss": 3.9705,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.137307442593931e-05,
+      "loss": 3.9613,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136470485723252e-05,
+      "loss": 3.9679,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1356318909722e-05,
+      "loss": 3.9591,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134793296221148e-05,
+      "loss": 3.9691,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133954701470096e-05,
+      "loss": 3.9712,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133116106719044e-05,
+      "loss": 3.9608,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132277511967992e-05,
+      "loss": 3.961,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131440555097313e-05,
+      "loss": 3.9707,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130601960346261e-05,
+      "loss": 3.9577,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129763365595209e-05,
+      "loss": 3.9588,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128924770844157e-05,
+      "loss": 3.968,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128086176093105e-05,
+      "loss": 3.9588,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127249219222426e-05,
+      "loss": 3.9744,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126410624471374e-05,
+      "loss": 3.9517,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.021423816680908,
+      "eval_runtime": 338.5827,
+      "eval_samples_per_second": 1127.024,
+      "eval_steps_per_second": 35.22,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125572029720322e-05,
+      "loss": 3.9227,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12473343496927e-05,
+      "loss": 3.9613,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123894840218218e-05,
+      "loss": 3.9524,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123056245467166e-05,
+      "loss": 3.9662,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122217650716114e-05,
+      "loss": 3.967,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121379055965062e-05,
+      "loss": 3.965,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12054046121401e-05,
+      "loss": 3.9567,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119701866462958e-05,
+      "loss": 3.9568,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118863271711906e-05,
+      "loss": 3.9563,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118024676960854e-05,
+      "loss": 3.9666,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117186082209802e-05,
+      "loss": 3.9627,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11634748745875e-05,
+      "loss": 3.9666,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115508892707698e-05,
+      "loss": 3.9628,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114670297956646e-05,
+      "loss": 3.9522,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113831703205594e-05,
+      "loss": 3.9577,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112993108454542e-05,
+      "loss": 3.948,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11215451370349e-05,
+      "loss": 3.9632,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111315918952438e-05,
+      "loss": 3.9581,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110477324201386e-05,
+      "loss": 3.9555,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109638729450334e-05,
+      "loss": 3.9578,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108801772579655e-05,
+      "loss": 3.9693,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1079648157089766e-05,
+      "loss": 3.9656,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1071262209579246e-05,
+      "loss": 3.9604,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1062876262068726e-05,
+      "loss": 3.9612,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1054490314558206e-05,
+      "loss": 3.9661,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1046104367047686e-05,
+      "loss": 3.9612,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1037718419537166e-05,
+      "loss": 3.9525,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102933247202664e-05,
+      "loss": 3.9619,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102094652451612e-05,
+      "loss": 3.9494,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1012576955809335e-05,
+      "loss": 3.9503,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1004191008298815e-05,
+      "loss": 3.9566,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099580506078829e-05,
+      "loss": 3.9575,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0987419113277774e-05,
+      "loss": 3.9652,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097904954457099e-05,
+      "loss": 3.9626,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097066359706047e-05,
+      "loss": 3.9599,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0962277649549943e-05,
+      "loss": 3.9661,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095390808084316e-05,
+      "loss": 3.9671,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094552213333264e-05,
+      "loss": 3.9474,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093713618582211e-05,
+      "loss": 3.9597,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092875023831159e-05,
+      "loss": 3.9477,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092036429080107e-05,
+      "loss": 3.953,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091199472209429e-05,
+      "loss": 3.9552,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090360877458376e-05,
+      "loss": 3.9626,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089522282707324e-05,
+      "loss": 3.9554,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088683687956273e-05,
+      "loss": 3.9526,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0878467310855944e-05,
+      "loss": 3.9615,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087008136334542e-05,
+      "loss": 3.9564,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08616954158349e-05,
+      "loss": 3.9532,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085330946832438e-05,
+      "loss": 3.9311,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0844939899617586e-05,
+      "loss": 3.9618,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0836553952107066e-05,
+      "loss": 3.9584,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0828168004596546e-05,
+      "loss": 3.9608,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0819782057086026e-05,
+      "loss": 3.9497,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811396109575506e-05,
+      "loss": 3.9442,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0803026540868715e-05,
+      "loss": 3.9478,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794640593358195e-05,
+      "loss": 3.9467,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078625464584768e-05,
+      "loss": 3.9609,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077786869833716e-05,
+      "loss": 3.9431,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076948275082664e-05,
+      "loss": 3.9613,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076109680331612e-05,
+      "loss": 3.952,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075272723460933e-05,
+      "loss": 3.9453,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074434128709881e-05,
+      "loss": 3.9578,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073595533958829e-05,
+      "loss": 3.9498,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072756939207777e-05,
+      "loss": 3.9484,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071918344456725e-05,
+      "loss": 3.9454,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071081387586046e-05,
+      "loss": 3.9608,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070242792834994e-05,
+      "loss": 3.9488,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069404198083942e-05,
+      "loss": 3.9538,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06856560333289e-05,
+      "loss": 3.9418,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067727008581838e-05,
+      "loss": 3.9488,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0668884138307866e-05,
+      "loss": 3.9539,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0660498190797346e-05,
+      "loss": 3.9672,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0652112243286826e-05,
+      "loss": 3.9505,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0643742674580035e-05,
+      "loss": 3.9577,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0635356727069515e-05,
+      "loss": 3.967,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0626987158362724e-05,
+      "loss": 3.9586,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0618601210852204e-05,
+      "loss": 3.9528,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061023164214541e-05,
+      "loss": 3.9581,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060184569463489e-05,
+      "loss": 3.9594,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059345974712437e-05,
+      "loss": 3.9534,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058507379961385e-05,
+      "loss": 3.9539,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057668785210333e-05,
+      "loss": 3.9648,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056830190459281e-05,
+      "loss": 3.968,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05599159570823e-05,
+      "loss": 3.9529,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055153000957178e-05,
+      "loss": 3.9384,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054314406206125e-05,
+      "loss": 3.9489,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053475811455073e-05,
+      "loss": 3.9598,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052637216704021e-05,
+      "loss": 3.9555,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051798621952969e-05,
+      "loss": 3.9608,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509616650822902e-05,
+      "loss": 3.9558,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501230703312382e-05,
+      "loss": 3.948,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0492844755801865e-05,
+      "loss": 3.9452,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0484458808291345e-05,
+      "loss": 3.9499,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0476072860780825e-05,
+      "loss": 3.9605,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0467686913270305e-05,
+      "loss": 3.9543,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459317344563514e-05,
+      "loss": 3.9717,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0450931397052994e-05,
+      "loss": 3.9495,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442545449542474e-05,
+      "loss": 3.9509,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434159502031957e-05,
+      "loss": 3.9516,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0425789933325167e-05,
+      "loss": 3.9587,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0417403985814647e-05,
+      "loss": 3.9612,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0409018038304126e-05,
+      "loss": 3.9527,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0400648469597336e-05,
+      "loss": 3.9578,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.039226252208682e-05,
+      "loss": 3.9453,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.03838765745763e-05,
+      "loss": 3.9599,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.037549062706578e-05,
+      "loss": 3.9388,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036710467955526e-05,
+      "loss": 3.9497,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035871873204474e-05,
+      "loss": 3.9528,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035033278453422e-05,
+      "loss": 3.9563,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.03419468370237e-05,
+      "loss": 3.9472,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033356088951318e-05,
+      "loss": 3.9535,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032517494200266e-05,
+      "loss": 3.9372,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0316788994492135e-05,
+      "loss": 3.9561,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308403046981615e-05,
+      "loss": 3.9489,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030003347827483e-05,
+      "loss": 3.9566,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.029164753076431e-05,
+      "loss": 3.9602,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0283261583253787e-05,
+      "loss": 3.9618,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0274875635743267e-05,
+      "loss": 3.9494,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0266489688232747e-05,
+      "loss": 3.951,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0258103740722227e-05,
+      "loss": 3.9539,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249734172015436e-05,
+      "loss": 3.9512,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241348224504916e-05,
+      "loss": 3.9547,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0232962276994396e-05,
+      "loss": 3.9491,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.022459270828761e-05,
+      "loss": 3.9537,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.021620676077709e-05,
+      "loss": 3.9452,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.020782081326657e-05,
+      "loss": 3.9471,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019943486575605e-05,
+      "loss": 3.9488,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019104891824553e-05,
+      "loss": 3.9496,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182679349538738e-05,
+      "loss": 3.9482,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.017429340202822e-05,
+      "loss": 3.9486,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01659074545177e-05,
+      "loss": 3.9597,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.015752150700718e-05,
+      "loss": 3.9642,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014913555949666e-05,
+      "loss": 3.9461,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014074961198614e-05,
+      "loss": 3.9547,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.013236366447562e-05,
+      "loss": 3.9602,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01239777169651e-05,
+      "loss": 3.9534,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.011559176945458e-05,
+      "loss": 3.9577,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0107222200747793e-05,
+      "loss": 3.949,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0098836253237273e-05,
+      "loss": 3.9599,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090466684530482e-05,
+      "loss": 3.9539,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0082080737019962e-05,
+      "loss": 3.9531,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073694789509442e-05,
+      "loss": 3.9506,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065308841998922e-05,
+      "loss": 3.9533,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0056922894488406e-05,
+      "loss": 3.9502,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0048553325781615e-05,
+      "loss": 3.9492,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0040167378271095e-05,
+      "loss": 3.9566,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031781430760575e-05,
+      "loss": 3.9486,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023395483250055e-05,
+      "loss": 3.9613,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015009535739534e-05,
+      "loss": 3.9401,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006623588229014e-05,
+      "loss": 3.9502,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.015658855438232,
+      "eval_runtime": 294.7591,
+      "eval_samples_per_second": 1294.586,
+      "eval_steps_per_second": 40.457,
+      "step": 1221120
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 5.054793715597333e+17,
+  "trial_name": null,
+  "trial_params": null
+}