diff --git "a/checkpoint-1831680/trainer_state.json" "b/checkpoint-1831680/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1831680/trainer_state.json"
@@ -0,0 +1,21679 @@
+{
+  "best_metric": 3.9886837005615234,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sim-ques/lstm/2/checkpoints/checkpoint-1831680",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1831680,
+  "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": 10.8183,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5489,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0575,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9912,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9451,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9054,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7488,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6422,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5422,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.463,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.394,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3437,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2742,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2007,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1518,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0921,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0583,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.003,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9592,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9277,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8867,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8583,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8315,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.8007,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7722,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7465,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.7163,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.6883,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6622,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6488,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6247,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 5.6034,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 5.5806,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 5.561,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 5.5533,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5428,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 5.5181,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 5.4789,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 5.4864,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.45,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4447,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4349,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4261,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.4039,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.3995,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3925,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3801,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3701,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 5.3509,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3372,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3406,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3213,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3104,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2914,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.296,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2761,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2905,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.262,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2589,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 5.2526,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2303,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2338,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2184,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2085,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2025,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2191,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 5.1953,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 5.1825,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 5.1629,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.1616,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.1757,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1713,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1457,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1612,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1531,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1385,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1181,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.123,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1065,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1038,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.1046,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.0941,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0968,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0916,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0698,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0726,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0717,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.065,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0531,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0506,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0492,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.036,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922872213228439e-05,
+      "loss": 5.0395,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0249,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0264,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.023,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0142,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9944,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9935,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.993,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 5.0033,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.99,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9787,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9698,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9755,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9503,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9518,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9484,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9505,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9401,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9479,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.926,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9473,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9266,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9303,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9301,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9286,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9137,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.904,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.9116,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9022,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.9033,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8983,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.886,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8983,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8775,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.873,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8678,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8738,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8581,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8763,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.867,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8559,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8472,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.8502,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.8479,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.8503,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.8473,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.8448,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.8353,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8306,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8197,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8183,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8236,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8049,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8185,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8176,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8064,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8111,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.806,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.7636847496032715,
+      "eval_runtime": 287.1664,
+      "eval_samples_per_second": 1328.815,
+      "eval_steps_per_second": 41.526,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7863,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.7912,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.8017,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7838,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7911,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.7784,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.7806,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.7683,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.7765,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.764,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7869,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7759,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7565,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7552,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7508,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7525,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.7462,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7448,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7411,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7515,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7425,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7463,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.7417,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7484,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7249,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7254,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7203,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.7187,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.7188,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7158,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7174,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7135,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7035,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7163,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.7168,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.7033,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.6867,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.7012,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.6791,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.6895,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6889,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.6926,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6803,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6868,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6885,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.682,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6828,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6663,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6679,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6763,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6651,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6669,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.6486,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.661,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6485,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.6664,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6502,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6555,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.656,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6334,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6454,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6441,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.6283,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214415574801015e-05,
+      "loss": 4.6323,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206029627290495e-05,
+      "loss": 4.6524,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197643679779975e-05,
+      "loss": 4.637,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.623,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6199,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.6232,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.6398,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.6394,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6222,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.6385,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6365,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.627,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6114,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6254,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.6135,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.6097,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.6136,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807190360353337e-05,
+      "loss": 4.6068,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806351765602285e-05,
+      "loss": 4.622,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805513170851233e-05,
+      "loss": 4.608,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804674576100181e-05,
+      "loss": 4.5992,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.6056,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.6053,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.6042,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.6,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004832402252944e-05,
+      "loss": 4.5936,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996446454742424e-05,
+      "loss": 4.6036,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79880605072319e-05,
+      "loss": 4.5906,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797967455972138e-05,
+      "loss": 4.5984,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797130499101459e-05,
+      "loss": 4.583,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796291904350407e-05,
+      "loss": 4.5969,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954533095993546e-05,
+      "loss": 4.5933,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946147148483026e-05,
+      "loss": 4.5866,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793777757977624e-05,
+      "loss": 4.5724,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5779,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5811,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5902,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5828,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.5724,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788747827351685e-05,
+      "loss": 4.5685,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787909232600633e-05,
+      "loss": 4.5772,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.551,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.564,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78539508622785e-05,
+      "loss": 4.5616,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784556491476798e-05,
+      "loss": 4.5647,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783717896725746e-05,
+      "loss": 4.557,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.5645,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820423451040155e-05,
+      "loss": 4.5532,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.5739,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.5579,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5558,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.5717,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.5686,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770124144780764e-05,
+      "loss": 4.5532,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761738197270244e-05,
+      "loss": 4.5459,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.56,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774498268105293e-05,
+      "loss": 4.5499,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773659673354241e-05,
+      "loss": 4.5496,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772821078603189e-05,
+      "loss": 4.557,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771982483852137e-05,
+      "loss": 4.5417,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771143889101086e-05,
+      "loss": 4.556,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770305294350034e-05,
+      "loss": 4.5384,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769466699598982e-05,
+      "loss": 4.541,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76862810484793e-05,
+      "loss": 4.5307,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767789510096878e-05,
+      "loss": 4.5426,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766950915345826e-05,
+      "loss": 4.5333,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766112320594773e-05,
+      "loss": 4.5483,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765273725843721e-05,
+      "loss": 4.5465,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764436768973043e-05,
+      "loss": 4.5361,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76359817422199e-05,
+      "loss": 4.522,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762759579470938e-05,
+      "loss": 4.5371,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.532,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5388,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760243795217783e-05,
+      "loss": 4.5353,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759406838347104e-05,
+      "loss": 4.5326,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758569881476425e-05,
+      "loss": 4.5292,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757731286725373e-05,
+      "loss": 4.5342,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568926919743206e-05,
+      "loss": 4.516,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560540972232686e-05,
+      "loss": 4.517,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552155024722166e-05,
+      "loss": 4.5246,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543785456015375e-05,
+      "loss": 4.5109,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.5199,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.5305,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5167,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510241665973295e-05,
+      "loss": 4.5154,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750185571846278e-05,
+      "loss": 4.5179,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.481045722961426,
+      "eval_runtime": 332.9874,
+      "eval_samples_per_second": 1145.962,
+      "eval_steps_per_second": 35.812,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749346977095226e-05,
+      "loss": 4.5024,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.506,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.52,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5059,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5133,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.4983,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.5113,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743476813837862e-05,
+      "loss": 4.4917,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426382190868095e-05,
+      "loss": 4.5049,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.4983,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.5128,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.5127,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4881,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.4937,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.4951,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736769693709819e-05,
+      "loss": 4.4896,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735931098958767e-05,
+      "loss": 4.4896,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4865,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.4846,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.5061,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4866,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.4988,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4938,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4998,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4727,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4842,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4736,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267098324579415e-05,
+      "loss": 4.481,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258712377068895e-05,
+      "loss": 4.4741,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250326429558375e-05,
+      "loss": 4.476,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.483,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4766,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4663,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216799018320024e-05,
+      "loss": 4.4829,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.477,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4731,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719167393339592e-05,
+      "loss": 4.4601,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.4713,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4514,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4664,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4598,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714976057464706e-05,
+      "loss": 4.4672,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4614,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4643,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4672,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.4632,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4666,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4484,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4481,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082689373366627e-05,
+      "loss": 4.4632,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.4508,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4518,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.4383,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704916196212828e-05,
+      "loss": 4.4514,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4367,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703240644591097e-05,
+      "loss": 4.4539,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702402049840045e-05,
+      "loss": 4.4426,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015634550889924e-05,
+      "loss": 4.4515,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4479,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4315,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4398,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.44,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4274,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4377,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.4455,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4406,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4274,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4208,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4272,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4427,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4469,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4306,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4431,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688149214832907e-05,
+      "loss": 4.4432,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687310620081855e-05,
+      "loss": 4.4396,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686472025330803e-05,
+      "loss": 4.4153,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.436,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4304,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.4186,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.4308,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4209,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.434,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4268,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4144,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.4227,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4212,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.4247,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676413801959298e-05,
+      "loss": 4.419,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.4151,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.4223,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4116,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4191,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672224103964785e-05,
+      "loss": 4.4084,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.4234,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705485523430536e-05,
+      "loss": 4.4185,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697099575920016e-05,
+      "loss": 4.4112,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688713628409496e-05,
+      "loss": 4.4026,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680327680898976e-05,
+      "loss": 4.4107,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.4041,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4221,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655169838367416e-05,
+      "loss": 4.4101,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4037,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.3977,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.4062,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.3859,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.3964,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3962,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.3994,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658813139469445e-05,
+      "loss": 4.3888,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657974544718393e-05,
+      "loss": 4.403,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657135949967341e-05,
+      "loss": 4.3877,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656297355216288e-05,
+      "loss": 4.4113,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655458760465237e-05,
+      "loss": 4.3956,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.3926,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.4121,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.4106,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652106019341402e-05,
+      "loss": 4.3946,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65126742459035e-05,
+      "loss": 4.3881,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.3965,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.3928,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648753278217567e-05,
+      "loss": 4.3896,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647914683466515e-05,
+      "loss": 4.4028,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.3825,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.4007,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.3853,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.3885,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.375,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.3923,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.3769,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.3977,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.3908,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3859,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.3701,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637854822214637e-05,
+      "loss": 4.3877,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637016227463585e-05,
+      "loss": 4.3847,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636177632712533e-05,
+      "loss": 4.3848,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635340675841854e-05,
+      "loss": 4.3823,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3889,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.3823,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.3889,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.3691,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.3727,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.3792,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62947051258449e-05,
+      "loss": 4.3657,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628631917833438e-05,
+      "loss": 4.3761,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.3862,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3708,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.3713,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.3786,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.346052169799805,
+      "eval_runtime": 305.7015,
+      "eval_samples_per_second": 1248.247,
+      "eval_steps_per_second": 39.009,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244422198389246e-05,
+      "loss": 4.3682,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236036250878726e-05,
+      "loss": 4.3637,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227650303368205e-05,
+      "loss": 4.3777,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3616,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210894787150895e-05,
+      "loss": 4.3755,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202508839640374e-05,
+      "loss": 4.3589,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3712,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3524,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.3678,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168981428402023e-05,
+      "loss": 4.3601,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.3764,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.3739,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614385634347793e-05,
+      "loss": 4.353,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135470395967415e-05,
+      "loss": 4.3614,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127084448456895e-05,
+      "loss": 4.3581,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611869850094637e-05,
+      "loss": 4.3577,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611031255343585e-05,
+      "loss": 4.3507,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610192660592533e-05,
+      "loss": 4.3534,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609354065841481e-05,
+      "loss": 4.3502,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608515471090429e-05,
+      "loss": 4.3772,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607676876339377e-05,
+      "loss": 4.3538,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606838281588325e-05,
+      "loss": 4.3601,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605999686837273e-05,
+      "loss": 4.3657,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605161092086221e-05,
+      "loss": 4.3675,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3435,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.3512,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3461,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018083509623864e-05,
+      "loss": 4.3523,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009697562113344e-05,
+      "loss": 4.3455,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001311614602824e-05,
+      "loss": 4.3442,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5992925667092303e-05,
+      "loss": 4.3535,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598455609838551e-05,
+      "loss": 4.3517,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597617015087499e-05,
+      "loss": 4.3462,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596778420336447e-05,
+      "loss": 4.3525,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595941463465768e-05,
+      "loss": 4.3528,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595102868714716e-05,
+      "loss": 4.3467,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3324,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3495,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3231,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3405,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3359,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3451,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589232705457352e-05,
+      "loss": 4.3448,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883941107063e-05,
+      "loss": 4.3378,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587557153835621e-05,
+      "loss": 4.3452,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586718559084569e-05,
+      "loss": 4.3448,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3442,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3269,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.3266,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3455,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3291,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3287,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808500337075786e-05,
+      "loss": 4.3235,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800114389565266e-05,
+      "loss": 4.3335,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791728442054746e-05,
+      "loss": 4.3211,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3329,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3275,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3318,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3318,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3194,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741429135795355e-05,
+      "loss": 4.321,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733043188284835e-05,
+      "loss": 4.3301,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724657240774315e-05,
+      "loss": 4.3116,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.3213,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570791810336074e-05,
+      "loss": 4.3292,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569953215585022e-05,
+      "loss": 4.3307,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56911462083397e-05,
+      "loss": 4.313,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568276026082918e-05,
+      "loss": 4.3101,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.3123,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3265,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565760241829762e-05,
+      "loss": 4.3334,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.3201,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3291,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3326,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3288,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561570543835248e-05,
+      "loss": 4.3079,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560733586964569e-05,
+      "loss": 4.325,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559894992213517e-05,
+      "loss": 4.3205,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559056397462465e-05,
+      "loss": 4.3098,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558217802711413e-05,
+      "loss": 4.3186,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557379207960361e-05,
+      "loss": 4.3141,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556540613209309e-05,
+      "loss": 4.3192,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557020184582566e-05,
+      "loss": 4.3196,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548634237072046e-05,
+      "loss": 4.3085,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.3174,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3116,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.315,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3127,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.3069,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498334930812655e-05,
+      "loss": 4.316,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548994898330214e-05,
+      "loss": 4.3107,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.308,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547319346708483e-05,
+      "loss": 4.3045,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546480751957431e-05,
+      "loss": 4.3195,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545642157206379e-05,
+      "loss": 4.313,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544803562455327e-05,
+      "loss": 4.3054,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.2982,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.3047,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2998,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541450821331492e-05,
+      "loss": 4.3217,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54061222658044e-05,
+      "loss": 4.3066,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539773631829388e-05,
+      "loss": 4.3018,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389366749587095e-05,
+      "loss": 4.2964,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380980802076575e-05,
+      "loss": 4.306,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372594854566055e-05,
+      "loss": 4.2791,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364208907055535e-05,
+      "loss": 4.2953,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355822959545015e-05,
+      "loss": 4.2957,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347437012034495e-05,
+      "loss": 4.3019,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339051064523975e-05,
+      "loss": 4.2868,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330665117013455e-05,
+      "loss": 4.303,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532227916950293e-05,
+      "loss": 4.2875,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531389322199241e-05,
+      "loss": 4.3057,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2987,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2913,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.3125,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280365810754064e-05,
+      "loss": 4.3121,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5271979863243544e-05,
+      "loss": 4.2995,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263593915733024e-05,
+      "loss": 4.2904,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255207968222504e-05,
+      "loss": 4.2972,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524683839951571e-05,
+      "loss": 4.2926,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523845245200519e-05,
+      "loss": 4.2917,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523006650449467e-05,
+      "loss": 4.3078,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2867,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.3019,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520490866196311e-05,
+      "loss": 4.2917,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519652271445259e-05,
+      "loss": 4.2966,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518813676694207e-05,
+      "loss": 4.2716,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517978357703902e-05,
+      "loss": 4.2985,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51713976295285e-05,
+      "loss": 4.2831,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516301168201798e-05,
+      "loss": 4.3,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515462573450746e-05,
+      "loss": 4.2922,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514623978699694e-05,
+      "loss": 4.2958,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137870218290147e-05,
+      "loss": 4.2753,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129484270779626e-05,
+      "loss": 4.2967,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121098323269106e-05,
+      "loss": 4.2872,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112712375758586e-05,
+      "loss": 4.2913,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104326428248066e-05,
+      "loss": 4.288,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095956859541275e-05,
+      "loss": 4.2979,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087570912030755e-05,
+      "loss": 4.2915,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079184964520235e-05,
+      "loss": 4.2925,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070799017009715e-05,
+      "loss": 4.2777,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.2822,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2838,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.2773,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.2847,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2919,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.2819,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501211376323981e-05,
+      "loss": 4.2789,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500372781572929e-05,
+      "loss": 4.2891,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.26270866394043,
+      "eval_runtime": 301.5381,
+      "eval_samples_per_second": 1265.482,
+      "eval_steps_per_second": 39.547,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995341868218764e-05,
+      "loss": 4.2721,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986955920708244e-05,
+      "loss": 4.2724,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978569973197724e-05,
+      "loss": 4.2855,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970184025687204e-05,
+      "loss": 4.2746,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961798078176684e-05,
+      "loss": 4.2882,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953412130666164e-05,
+      "loss": 4.2707,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494502618315565e-05,
+      "loss": 4.2776,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493664023564513e-05,
+      "loss": 4.2641,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492825428813461e-05,
+      "loss": 4.2831,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491986834062409e-05,
+      "loss": 4.2719,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491148239311357e-05,
+      "loss": 4.2828,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490309644560305e-05,
+      "loss": 4.2879,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489472687689626e-05,
+      "loss": 4.2614,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488634092938574e-05,
+      "loss": 4.2791,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487795498187522e-05,
+      "loss": 4.2697,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48695690343647e-05,
+      "loss": 4.2704,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486118308685418e-05,
+      "loss": 4.2604,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485281351814739e-05,
+      "loss": 4.2688,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484442757063687e-05,
+      "loss": 4.2658,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483604162312635e-05,
+      "loss": 4.2883,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482765567561583e-05,
+      "loss": 4.267,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481926972810531e-05,
+      "loss": 4.2755,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481088378059479e-05,
+      "loss": 4.2801,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480249783308427e-05,
+      "loss": 4.283,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479411188557375e-05,
+      "loss": 4.2589,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478572593806323e-05,
+      "loss": 4.2706,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477735636935644e-05,
+      "loss": 4.2583,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476897042184592e-05,
+      "loss": 4.2676,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47605844743354e-05,
+      "loss": 4.2584,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752214905628606e-05,
+      "loss": 4.2632,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743828958118086e-05,
+      "loss": 4.2671,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473544301060757e-05,
+      "loss": 4.2671,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472705706309705e-05,
+      "loss": 4.2611,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471867111558653e-05,
+      "loss": 4.2722,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471028516807601e-05,
+      "loss": 4.2644,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470189922056549e-05,
+      "loss": 4.2688,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469351327305497e-05,
+      "loss": 4.2524,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468514370434818e-05,
+      "loss": 4.2653,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467675775683766e-05,
+      "loss": 4.2484,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466837180932714e-05,
+      "loss": 4.2556,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465998586181662e-05,
+      "loss": 4.2548,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46515999143061e-05,
+      "loss": 4.2641,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464321396679558e-05,
+      "loss": 4.2616,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463482801928506e-05,
+      "loss": 4.2597,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462644207177454e-05,
+      "loss": 4.2626,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461807250306775e-05,
+      "loss": 4.2641,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460968655555724e-05,
+      "loss": 4.2593,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460130060804672e-05,
+      "loss": 4.2528,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459291466053619e-05,
+      "loss": 4.2403,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458452871302567e-05,
+      "loss": 4.2637,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576159144318886e-05,
+      "loss": 4.2536,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567789575612095e-05,
+      "loss": 4.2472,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559403628101575e-05,
+      "loss": 4.2476,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551017680591055e-05,
+      "loss": 4.2538,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542631733080535e-05,
+      "loss": 4.2448,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453424578557001e-05,
+      "loss": 4.247,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452585983805949e-05,
+      "loss": 4.2502,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517473890548975e-05,
+      "loss": 4.2482,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509087943038455e-05,
+      "loss": 4.2581,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500718374331664e-05,
+      "loss": 4.2394,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492332426821144e-05,
+      "loss": 4.2432,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448396285811436e-05,
+      "loss": 4.2536,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447557691060384e-05,
+      "loss": 4.2375,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446719096309331e-05,
+      "loss": 4.2442,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445880501558279e-05,
+      "loss": 4.2492,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445041906807227e-05,
+      "loss": 4.2521,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444203312056175e-05,
+      "loss": 4.2383,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443366355185496e-05,
+      "loss": 4.2325,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442527760434444e-05,
+      "loss": 4.2301,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441689165683393e-05,
+      "loss": 4.2476,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440850570932341e-05,
+      "loss": 4.2591,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440011976181289e-05,
+      "loss": 4.2423,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439173381430237e-05,
+      "loss": 4.2542,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438334786679185e-05,
+      "loss": 4.2553,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437496191928133e-05,
+      "loss": 4.2568,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436659235057454e-05,
+      "loss": 4.2324,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435820640306402e-05,
+      "loss": 4.2472,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349836834357227e-05,
+      "loss": 4.2448,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341450886846707e-05,
+      "loss": 4.2331,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333064939336186e-05,
+      "loss": 4.2497,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324678991825666e-05,
+      "loss": 4.2385,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316293044315146e-05,
+      "loss": 4.2413,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307907096804626e-05,
+      "loss": 4.2457,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429953752809784e-05,
+      "loss": 4.2393,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429115158058732e-05,
+      "loss": 4.2442,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42827656330768e-05,
+      "loss": 4.2353,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427439606437001e-05,
+      "loss": 4.2416,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426601011685949e-05,
+      "loss": 4.2343,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425762416934897e-05,
+      "loss": 4.2353,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424923822183845e-05,
+      "loss": 4.2407,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424085227432793e-05,
+      "loss": 4.2396,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423246632681741e-05,
+      "loss": 4.2327,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422408037930689e-05,
+      "loss": 4.2324,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421569443179637e-05,
+      "loss": 4.2476,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420730848428585e-05,
+      "loss": 4.2419,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419892253677533e-05,
+      "loss": 4.2359,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419053658926481e-05,
+      "loss": 4.2277,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418216702055803e-05,
+      "loss": 4.2305,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41737810730475e-05,
+      "loss": 4.2296,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416539512553698e-05,
+      "loss": 4.2438,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415700917802646e-05,
+      "loss": 4.2416,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148639609319676e-05,
+      "loss": 4.2305,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414025366180915e-05,
+      "loss": 4.2245,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4131884093102365e-05,
+      "loss": 4.238,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123498145591845e-05,
+      "loss": 4.2051,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411511219808132e-05,
+      "loss": 4.2295,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106726250570805e-05,
+      "loss": 4.219,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409835668186402e-05,
+      "loss": 4.2314,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899707343535e-05,
+      "loss": 4.2173,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081584786842974e-05,
+      "loss": 4.2301,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073198839332453e-05,
+      "loss": 4.2248,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064812891821933e-05,
+      "loss": 4.2285,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056426944311413e-05,
+      "loss": 4.2336,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404804099680089e-05,
+      "loss": 4.2191,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40396714280941e-05,
+      "loss": 4.243,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403128548058358e-05,
+      "loss": 4.2447,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402289953307306e-05,
+      "loss": 4.231,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401451358556254e-05,
+      "loss": 4.2186,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400612763805202e-05,
+      "loss": 4.2307,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39977416905415e-05,
+      "loss": 4.2265,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398935574303098e-05,
+      "loss": 4.2191,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39809861743242e-05,
+      "loss": 4.2387,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397260022681368e-05,
+      "loss": 4.2186,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396421427930316e-05,
+      "loss": 4.2312,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395582833179264e-05,
+      "loss": 4.2254,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394744238428212e-05,
+      "loss": 4.2265,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393907281557533e-05,
+      "loss": 4.2137,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393068686806481e-05,
+      "loss": 4.2233,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392230092055429e-05,
+      "loss": 4.2208,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391391497304377e-05,
+      "loss": 4.2344,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390552902553325e-05,
+      "loss": 4.2243,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389714307802273e-05,
+      "loss": 4.2267,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388875713051221e-05,
+      "loss": 4.2102,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388038756180542e-05,
+      "loss": 4.2312,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387201799309863e-05,
+      "loss": 4.2232,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386363204558811e-05,
+      "loss": 4.2274,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385524609807759e-05,
+      "loss": 4.2208,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384686015056707e-05,
+      "loss": 4.2317,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383847420305655e-05,
+      "loss": 4.2214,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383008825554603e-05,
+      "loss": 4.2314,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382171868683924e-05,
+      "loss": 4.21,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381333273932872e-05,
+      "loss": 4.2191,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38049467918182e-05,
+      "loss": 4.2171,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379656084430768e-05,
+      "loss": 4.2126,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378817489679716e-05,
+      "loss": 4.2225,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377978894928664e-05,
+      "loss": 4.2226,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771419380579856e-05,
+      "loss": 4.2213,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763033433069336e-05,
+      "loss": 4.2083,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375464748555881e-05,
+      "loss": 4.2257,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.205767631530762,
+      "eval_runtime": 294.8186,
+      "eval_samples_per_second": 1294.325,
+      "eval_steps_per_second": 40.449,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.2148,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3737891969341505e-05,
+      "loss": 4.206,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372950602183098e-05,
+      "loss": 4.2234,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372112007432046e-05,
+      "loss": 4.2132,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371273412680994e-05,
+      "loss": 4.2173,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370434817929942e-05,
+      "loss": 4.2116,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36959622317889e-05,
+      "loss": 4.2111,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368757628427838e-05,
+      "loss": 4.2032,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367919033676786e-05,
+      "loss": 4.2167,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670820768061074e-05,
+      "loss": 4.2114,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662434820550554e-05,
+      "loss": 4.2186,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654048873040034e-05,
+      "loss": 4.2232,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364567930433324e-05,
+      "loss": 4.2013,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363729335682272e-05,
+      "loss": 4.2156,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36289074093122e-05,
+      "loss": 4.2112,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362052146180168e-05,
+      "loss": 4.2049,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361213551429116e-05,
+      "loss": 4.1996,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360374956678064e-05,
+      "loss": 4.2073,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359536361927012e-05,
+      "loss": 4.2022,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35869776717596e-05,
+      "loss": 4.2223,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357859172424908e-05,
+      "loss": 4.2114,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2129,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356186896563924e-05,
+      "loss": 4.2203,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355348301812872e-05,
+      "loss": 4.222,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35450970706182e-05,
+      "loss": 4.1949,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3536711123107677e-05,
+      "loss": 4.2098,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528325175597157e-05,
+      "loss": 4.2001,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3519939228086636e-05,
+      "loss": 4.2051,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511569659379846e-05,
+      "loss": 4.1987,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503183711869326e-05,
+      "loss": 4.2028,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494797764358805e-05,
+      "loss": 4.2088,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486411816848285e-05,
+      "loss": 4.2053,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478025869337765e-05,
+      "loss": 4.2032,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346963992182725e-05,
+      "loss": 4.2065,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346125397431673e-05,
+      "loss": 4.2046,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345286802680621e-05,
+      "loss": 4.2068,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344448207929569e-05,
+      "loss": 4.1961,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436096131785165e-05,
+      "loss": 4.1967,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427710184274645e-05,
+      "loss": 4.193,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419324236764125e-05,
+      "loss": 4.1936,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3410938289253605e-05,
+      "loss": 4.1959,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402552341743085e-05,
+      "loss": 4.2039,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394166394232565e-05,
+      "loss": 4.2021,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385796825525774e-05,
+      "loss": 4.2006,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377410878015254e-05,
+      "loss": 4.2088,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369024930504734e-05,
+      "loss": 4.2011,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360638982994214e-05,
+      "loss": 4.2032,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335226941428743e-05,
+      "loss": 4.1935,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334388346677691e-05,
+      "loss": 4.1804,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333549751926639e-05,
+      "loss": 4.2082,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332711157175587e-05,
+      "loss": 4.1928,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331874200304908e-05,
+      "loss": 4.1941,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331035605553856e-05,
+      "loss": 4.1909,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330197010802804e-05,
+      "loss": 4.192,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329358416051752e-05,
+      "loss": 4.186,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285198213007e-05,
+      "loss": 4.1883,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327681226549648e-05,
+      "loss": 4.1932,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326844269678969e-05,
+      "loss": 4.1911,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260073128082904e-05,
+      "loss": 4.2012,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3251687180572383e-05,
+      "loss": 4.1819,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243301233061863e-05,
+      "loss": 4.1861,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323491528555134e-05,
+      "loss": 4.1984,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322652933804082e-05,
+      "loss": 4.181,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32181433905303e-05,
+      "loss": 4.1847,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320975744301978e-05,
+      "loss": 4.1886,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320137149550926e-05,
+      "loss": 4.1985,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319298554799874e-05,
+      "loss": 4.182,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1813,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317623003178143e-05,
+      "loss": 4.1704,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316784408427091e-05,
+      "loss": 4.1934,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315945813676039e-05,
+      "loss": 4.2024,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315107218924987e-05,
+      "loss": 4.1895,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314268624173935e-05,
+      "loss": 4.1948,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313431667303257e-05,
+      "loss": 4.1985,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.2041,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311756115681526e-05,
+      "loss": 4.1773,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310917520930474e-05,
+      "loss": 4.1923,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310078926179422e-05,
+      "loss": 4.1895,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30924033142837e-05,
+      "loss": 4.1751,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308401736677318e-05,
+      "loss": 4.1928,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307563141926265e-05,
+      "loss": 4.1886,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306724547175213e-05,
+      "loss": 4.1859,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305885952424161e-05,
+      "loss": 4.1918,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050489955534826e-05,
+      "loss": 4.1841,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042104008024306e-05,
+      "loss": 4.1809,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033718060513786e-05,
+      "loss": 4.1846,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025332113003266e-05,
+      "loss": 4.191,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016946165492746e-05,
+      "loss": 4.176,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008576596785955e-05,
+      "loss": 4.1831,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000190649275435e-05,
+      "loss": 4.1892,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991804701764915e-05,
+      "loss": 4.1851,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983418754254395e-05,
+      "loss": 4.1796,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975032806743874e-05,
+      "loss": 4.1783,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966646859233354e-05,
+      "loss": 4.1942,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.1873,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.1844,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294150539550553e-05,
+      "loss": 4.1694,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293311944799501e-05,
+      "loss": 4.1777,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1779,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29163639317777e-05,
+      "loss": 4.1879,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290797798426718e-05,
+      "loss": 4.1895,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289959203675666e-05,
+      "loss": 4.1763,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289120608924614e-05,
+      "loss": 4.1714,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288282014173562e-05,
+      "loss": 4.1834,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287445057302883e-05,
+      "loss": 4.155,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286606462551831e-05,
+      "loss": 4.1771,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285767867800779e-05,
+      "loss": 4.1626,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284929273049727e-05,
+      "loss": 4.1833,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284092316179048e-05,
+      "loss": 4.1645,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2832537214279964e-05,
+      "loss": 4.1801,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824151266769444e-05,
+      "loss": 4.1698,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2815765319258924e-05,
+      "loss": 4.1754,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807379371748404e-05,
+      "loss": 4.1813,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279900980304161e-05,
+      "loss": 4.1726,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279062385553109e-05,
+      "loss": 4.1878,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278223790802057e-05,
+      "loss": 4.1911,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277385196051005e-05,
+      "loss": 4.1804,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276546601299953e-05,
+      "loss": 4.1687,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275708006548901e-05,
+      "loss": 4.1721,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748694117978486e-05,
+      "loss": 4.1767,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740308170467966e-05,
+      "loss": 4.1647,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2731922222957446e-05,
+      "loss": 4.1919,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1628,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271518308554388e-05,
+      "loss": 4.1825,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270679713803336e-05,
+      "loss": 4.1739,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698427569326566e-05,
+      "loss": 4.1735,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690041621816046e-05,
+      "loss": 4.1629,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681672053109256e-05,
+      "loss": 4.174,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673286105598735e-05,
+      "loss": 4.1691,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664900158088215e-05,
+      "loss": 4.1828,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656514210577695e-05,
+      "loss": 4.174,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648128263067175e-05,
+      "loss": 4.1765,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639742315556655e-05,
+      "loss": 4.1583,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631356368046135e-05,
+      "loss": 4.1808,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2622970420535615e-05,
+      "loss": 4.171,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614584473025095e-05,
+      "loss": 4.1752,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606198525514575e-05,
+      "loss": 4.1703,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597812578004055e-05,
+      "loss": 4.1794,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589426630493535e-05,
+      "loss": 4.1699,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581057061786744e-05,
+      "loss": 4.1842,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572671114276224e-05,
+      "loss": 4.1579,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564285166765704e-05,
+      "loss": 4.1692,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2555899219255184e-05,
+      "loss": 4.1678,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547513271744664e-05,
+      "loss": 4.1652,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539127324234144e-05,
+      "loss": 4.1661,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530741376723624e-05,
+      "loss": 4.1734,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522355429213104e-05,
+      "loss": 4.175,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251398586050632e-05,
+      "loss": 4.1585,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250561629179953e-05,
+      "loss": 4.1775,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.165190696716309,
+      "eval_runtime": 294.1315,
+      "eval_samples_per_second": 1297.348,
+      "eval_steps_per_second": 40.543,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249723034428901e-05,
+      "loss": 4.1667,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248884439677849e-05,
+      "loss": 4.1534,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248045844926797e-05,
+      "loss": 4.1751,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247207250175745e-05,
+      "loss": 4.1639,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246368655424693e-05,
+      "loss": 4.169,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245530060673641e-05,
+      "loss": 4.1609,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244691465922589e-05,
+      "loss": 4.1595,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243852871171537e-05,
+      "loss": 4.1592,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243014276420485e-05,
+      "loss": 4.1655,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242175681669432e-05,
+      "loss": 4.1619,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241337086918381e-05,
+      "loss": 4.1725,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240498492167329e-05,
+      "loss": 4.1737,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23966153529665e-05,
+      "loss": 4.1581,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238822940545598e-05,
+      "loss": 4.1644,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237984345794546e-05,
+      "loss": 4.1654,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237145751043494e-05,
+      "loss": 4.1545,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363087941728146e-05,
+      "loss": 4.1511,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354701994217626e-05,
+      "loss": 4.1587,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346316046707106e-05,
+      "loss": 4.1572,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337930099196586e-05,
+      "loss": 4.17,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329544151686066e-05,
+      "loss": 4.167,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321158204175546e-05,
+      "loss": 4.1679,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312772256665026e-05,
+      "loss": 4.1692,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304386309154506e-05,
+      "loss": 4.1712,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296000361643986e-05,
+      "loss": 4.1518,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228761441413347e-05,
+      "loss": 4.164,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227922846662295e-05,
+      "loss": 4.1475,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227084251911243e-05,
+      "loss": 4.1568,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226248932920937e-05,
+      "loss": 4.156,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225410338169885e-05,
+      "loss": 4.1539,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224571743418833e-05,
+      "loss": 4.1624,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223733148667781e-05,
+      "loss": 4.1587,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222894553916729e-05,
+      "loss": 4.1572,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222055959165677e-05,
+      "loss": 4.159,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221219002294998e-05,
+      "loss": 4.1583,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220380407543946e-05,
+      "loss": 4.1618,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219541812792894e-05,
+      "loss": 4.1518,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187032180418426e-05,
+      "loss": 4.1519,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178646232907906e-05,
+      "loss": 4.1464,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170276664201115e-05,
+      "loss": 4.1472,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2161890716690595e-05,
+      "loss": 4.1517,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153504769180075e-05,
+      "loss": 4.1574,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145118821669555e-05,
+      "loss": 4.1558,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136732874159035e-05,
+      "loss": 4.155,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212834692664851e-05,
+      "loss": 4.1633,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2119977357941724e-05,
+      "loss": 4.1567,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111591410431204e-05,
+      "loss": 4.1596,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210320546292068e-05,
+      "loss": 4.1458,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094819515410164e-05,
+      "loss": 4.1343,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086433567899644e-05,
+      "loss": 4.1652,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207806399919286e-05,
+      "loss": 4.1477,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206967805168233e-05,
+      "loss": 4.1444,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206129210417181e-05,
+      "loss": 4.1501,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205290615666129e-05,
+      "loss": 4.1438,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204452020915077e-05,
+      "loss": 4.1463,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203615064044398e-05,
+      "loss": 4.1393,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202776469293346e-05,
+      "loss": 4.1479,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201937874542294e-05,
+      "loss": 4.1475,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201099279791242e-05,
+      "loss": 4.1602,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20026068504019e-05,
+      "loss": 4.1367,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199422090289138e-05,
+      "loss": 4.1377,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198583495538086e-05,
+      "loss": 4.1595,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197744900787035e-05,
+      "loss": 4.1319,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196907943916356e-05,
+      "loss": 4.1399,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196070987045677e-05,
+      "loss": 4.1476,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195232392294625e-05,
+      "loss": 4.1589,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194393797543573e-05,
+      "loss": 4.135,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193555202792521e-05,
+      "loss": 4.1391,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927182459218416e-05,
+      "loss": 4.1263,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918796511707896e-05,
+      "loss": 4.1453,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910410564197376e-05,
+      "loss": 4.1607,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902024616686856e-05,
+      "loss": 4.1464,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189365504798007e-05,
+      "loss": 4.146,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188526910046955e-05,
+      "loss": 4.1553,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187688315295903e-05,
+      "loss": 4.159,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186849720544851e-05,
+      "loss": 4.1374,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186011125793799e-05,
+      "loss": 4.1467,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185172531042747e-05,
+      "loss": 4.1474,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184333936291695e-05,
+      "loss": 4.1324,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183496979421016e-05,
+      "loss": 4.1465,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182658384669964e-05,
+      "loss": 4.1456,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181819789918912e-05,
+      "loss": 4.1452,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18098119516786e-05,
+      "loss": 4.1458,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180142600416808e-05,
+      "loss": 4.1462,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179304005665756e-05,
+      "loss": 4.1365,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178465410914704e-05,
+      "loss": 4.1453,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177626816163652e-05,
+      "loss": 4.1443,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767898592929736e-05,
+      "loss": 4.1381,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759529024222945e-05,
+      "loss": 4.1406,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751143076712425e-05,
+      "loss": 4.1386,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1742757129201905e-05,
+      "loss": 4.1455,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734371181691385e-05,
+      "loss": 4.1385,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725985234180865e-05,
+      "loss": 4.133,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171759928667034e-05,
+      "loss": 4.1548,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170924609676728e-05,
+      "loss": 4.1441,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170086014925676e-05,
+      "loss": 4.143,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169247420174624e-05,
+      "loss": 4.1282,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168408825423572e-05,
+      "loss": 4.1389,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16757023067252e-05,
+      "loss": 4.1345,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166731635921469e-05,
+      "loss": 4.1446,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165893041170417e-05,
+      "loss": 4.1521,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165054446419364e-05,
+      "loss": 4.1337,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164215851668312e-05,
+      "loss": 4.1339,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16337725691726e-05,
+      "loss": 4.1385,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162540300046581e-05,
+      "loss": 4.1148,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161701705295529e-05,
+      "loss": 4.1382,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160864748424851e-05,
+      "loss": 4.1196,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160026153673799e-05,
+      "loss": 4.1398,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159187558922746e-05,
+      "loss": 4.1253,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158348964171694e-05,
+      "loss": 4.1375,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157510369420643e-05,
+      "loss": 4.1306,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156671774669591e-05,
+      "loss": 4.1314,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155833179918539e-05,
+      "loss": 4.1462,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154994585167487e-05,
+      "loss": 4.1278,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154155990416435e-05,
+      "loss": 4.1474,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153317395665383e-05,
+      "loss": 4.1492,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524804387947036e-05,
+      "loss": 4.1372,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516418440436516e-05,
+      "loss": 4.1348,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508032492925996e-05,
+      "loss": 4.126,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499646545415476e-05,
+      "loss": 4.1351,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491260597904956e-05,
+      "loss": 4.1298,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482891029198165e-05,
+      "loss": 4.1491,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474505081687645e-05,
+      "loss": 4.1197,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466119134177125e-05,
+      "loss": 4.1419,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145773318666661e-05,
+      "loss": 4.1373,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144936361795982e-05,
+      "loss": 4.1324,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14409776704493e-05,
+      "loss": 4.122,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143259172293878e-05,
+      "loss": 4.1362,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142420577542826e-05,
+      "loss": 4.124,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141581982791774e-05,
+      "loss": 4.1478,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140743388040722e-05,
+      "loss": 4.1302,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13990479328967e-05,
+      "loss": 4.1379,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390661985386174e-05,
+      "loss": 4.1186,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138229241667939e-05,
+      "loss": 4.1439,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137390646916887e-05,
+      "loss": 4.1339,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136552052165835e-05,
+      "loss": 4.1375,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135713457414783e-05,
+      "loss": 4.1322,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348765005441045e-05,
+      "loss": 4.134,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340395436734254e-05,
+      "loss": 4.1341,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332009489223734e-05,
+      "loss": 4.149,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323623541713214e-05,
+      "loss": 4.1171,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315237594202694e-05,
+      "loss": 4.1306,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306851646692174e-05,
+      "loss": 4.127,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129846569918165e-05,
+      "loss": 4.1289,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129007975167113e-05,
+      "loss": 4.1256,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128169380416061e-05,
+      "loss": 4.1328,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127330785665009e-05,
+      "loss": 4.134,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649382879433e-05,
+      "loss": 4.1239,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125656871923652e-05,
+      "loss": 4.1404,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.134054660797119,
+      "eval_runtime": 297.5548,
+      "eval_samples_per_second": 1282.423,
+      "eval_steps_per_second": 40.077,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248182771726e-05,
+      "loss": 4.1255,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123979682421548e-05,
+      "loss": 4.1173,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123141087670495e-05,
+      "loss": 4.1353,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122302492919443e-05,
+      "loss": 4.1214,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121463898168391e-05,
+      "loss": 4.1299,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120625303417339e-05,
+      "loss": 4.1249,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119786708666287e-05,
+      "loss": 4.1249,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118948113915235e-05,
+      "loss": 4.1176,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118111157044556e-05,
+      "loss": 4.1294,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117272562293504e-05,
+      "loss": 4.121,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116433967542452e-05,
+      "loss": 4.1331,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1155953727914e-05,
+      "loss": 4.135,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114758415920722e-05,
+      "loss": 4.1241,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11391982116967e-05,
+      "loss": 4.1261,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113081226418618e-05,
+      "loss": 4.122,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112242631667566e-05,
+      "loss": 4.1161,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114056747968866e-05,
+      "loss": 4.1151,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105670800458346e-05,
+      "loss": 4.1199,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097284852947826e-05,
+      "loss": 4.1237,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1088898905437306e-05,
+      "loss": 4.1288,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1080512957926786e-05,
+      "loss": 4.1337,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072127010416266e-05,
+      "loss": 4.1294,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063741062905746e-05,
+      "loss": 4.1289,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055355115395225e-05,
+      "loss": 4.1345,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1046969167884705e-05,
+      "loss": 4.1106,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038583220374185e-05,
+      "loss": 4.1277,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10302136516674e-05,
+      "loss": 4.1153,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102182770415688e-05,
+      "loss": 4.1179,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101344175664636e-05,
+      "loss": 4.1166,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005055809135834e-05,
+      "loss": 4.1212,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996669861625314e-05,
+      "loss": 4.1219,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988283914114794e-05,
+      "loss": 4.1235,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097991434540801e-05,
+      "loss": 4.1211,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097152839789748e-05,
+      "loss": 4.1257,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096314245038696e-05,
+      "loss": 4.1144,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095475650287644e-05,
+      "loss": 4.1267,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094638693416966e-05,
+      "loss": 4.1217,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093800098665914e-05,
+      "loss": 4.1064,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092961503914862e-05,
+      "loss": 4.1102,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09212290916381e-05,
+      "loss": 4.1113,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091284314412758e-05,
+      "loss": 4.1184,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090445719661706e-05,
+      "loss": 4.1184,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089607124910654e-05,
+      "loss": 4.1187,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088770168039975e-05,
+      "loss": 4.1163,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087931573288923e-05,
+      "loss": 4.1289,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087092978537871e-05,
+      "loss": 4.1199,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086254383786819e-05,
+      "loss": 4.125,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085415789035767e-05,
+      "loss": 4.1125,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084578832165088e-05,
+      "loss": 4.0921,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083740237414036e-05,
+      "loss": 4.1307,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829016426629844e-05,
+      "loss": 4.1148,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082064685792305e-05,
+      "loss": 4.1117,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081226091041253e-05,
+      "loss": 4.1125,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080387496290201e-05,
+      "loss": 4.1072,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079548901539149e-05,
+      "loss": 4.1111,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078710306788097e-05,
+      "loss": 4.103,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077871712037045e-05,
+      "loss": 4.113,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077033117285993e-05,
+      "loss": 4.1084,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076194522534941e-05,
+      "loss": 4.1256,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075355927783889e-05,
+      "loss": 4.1021,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074517333032837e-05,
+      "loss": 4.1017,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0736787382817845e-05,
+      "loss": 4.1216,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728401435307325e-05,
+      "loss": 4.0957,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072003186660054e-05,
+      "loss": 4.1009,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071166229789376e-05,
+      "loss": 4.1146,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070327635038324e-05,
+      "loss": 4.1209,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069489040287272e-05,
+      "loss": 4.1021,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06865044553622e-05,
+      "loss": 4.1027,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067811850785167e-05,
+      "loss": 4.0927,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066973256034115e-05,
+      "loss": 4.1099,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661362991634366e-05,
+      "loss": 4.1258,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0652977044123846e-05,
+      "loss": 4.1104,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064459109661332e-05,
+      "loss": 4.1118,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06362051491028e-05,
+      "loss": 4.1214,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062781920159228e-05,
+      "loss": 4.1246,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619449632885495e-05,
+      "loss": 4.0988,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611063685374975e-05,
+      "loss": 4.1159,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602677737864455e-05,
+      "loss": 4.1078,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594291790353935e-05,
+      "loss": 4.1041,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585922221647144e-05,
+      "loss": 4.1083,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577536274136624e-05,
+      "loss": 4.1144,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0569150326626104e-05,
+      "loss": 4.1082,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0560764379115584e-05,
+      "loss": 4.1157,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0552378431605064e-05,
+      "loss": 4.1093,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0543992484094544e-05,
+      "loss": 4.0993,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0535606536584024e-05,
+      "loss": 4.1107,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0527220589073504e-05,
+      "loss": 4.1078,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051885102036671e-05,
+      "loss": 4.1036,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05104650728562e-05,
+      "loss": 4.107,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050209550414941e-05,
+      "loss": 4.1065,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049370955663889e-05,
+      "loss": 4.1118,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048532360912837e-05,
+      "loss": 4.1061,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047693766161785e-05,
+      "loss": 4.097,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046855171410733e-05,
+      "loss": 4.1173,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046018214540054e-05,
+      "loss": 4.1123,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045179619789002e-05,
+      "loss": 4.1087,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04434102503795e-05,
+      "loss": 4.0981,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043502430286898e-05,
+      "loss": 4.1013,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426654734162186e-05,
+      "loss": 4.0991,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418268786651666e-05,
+      "loss": 4.1086,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040988283914115e-05,
+      "loss": 4.115,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040149689163063e-05,
+      "loss": 4.103,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039311094412011e-05,
+      "loss": 4.1009,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038472499660959e-05,
+      "loss": 4.1001,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037633904909907e-05,
+      "loss": 4.0854,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036795310158855e-05,
+      "loss": 4.101,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035958353288176e-05,
+      "loss": 4.0889,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035121396417497e-05,
+      "loss": 4.1088,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034282801666445e-05,
+      "loss": 4.0927,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033444206915393e-05,
+      "loss": 4.1028,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032605612164341e-05,
+      "loss": 4.0947,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031767017413289e-05,
+      "loss": 4.1046,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030928422662237e-05,
+      "loss": 4.1089,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030089827911185e-05,
+      "loss": 4.0966,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029251233160133e-05,
+      "loss": 4.1144,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284142762894547e-05,
+      "loss": 4.1127,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275756815384027e-05,
+      "loss": 4.1023,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267370867873506e-05,
+      "loss": 4.105,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025898492036298e-05,
+      "loss": 4.0922,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025059897285246e-05,
+      "loss": 4.1044,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242229404145675e-05,
+      "loss": 4.1006,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0233859835438885e-05,
+      "loss": 4.1093,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0225473887928365e-05,
+      "loss": 4.0913,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217087940417845e-05,
+      "loss": 4.1093,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0208701992907324e-05,
+      "loss": 4.1068,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200316045396804e-05,
+      "loss": 4.1,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0191930097886284e-05,
+      "loss": 4.0867,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01835605291795e-05,
+      "loss": 4.1048,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017519096047271e-05,
+      "loss": 4.0936,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016680501296219e-05,
+      "loss": 4.1122,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015841906545167e-05,
+      "loss": 4.1018,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015003311794115e-05,
+      "loss": 4.1027,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014164717043063e-05,
+      "loss": 4.085,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01332612229201e-05,
+      "loss": 4.1151,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012487527540958e-05,
+      "loss": 4.0972,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011648932789906e-05,
+      "loss": 4.1046,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010810338038854e-05,
+      "loss": 4.0992,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009971743287803e-05,
+      "loss": 4.0999,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009133148536751e-05,
+      "loss": 4.104,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008294553785699e-05,
+      "loss": 4.1141,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00745759691502e-05,
+      "loss": 4.0865,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006619002163968e-05,
+      "loss": 4.1014,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005780407412916e-05,
+      "loss": 4.0973,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004941812661864e-05,
+      "loss": 4.0949,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004103217910812e-05,
+      "loss": 4.0938,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003266261040133e-05,
+      "loss": 4.1029,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002427666289081e-05,
+      "loss": 4.105,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0015907094184016e-05,
+      "loss": 4.0901,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0007521146673496e-05,
+      "loss": 4.1056,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.109273433685303,
+      "eval_runtime": 299.0586,
+      "eval_samples_per_second": 1275.974,
+      "eval_steps_per_second": 39.875,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999913519916298e-05,
+      "loss": 4.0921,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999074925165246e-05,
+      "loss": 4.0898,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998236330414194e-05,
+      "loss": 4.1006,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997397735663142e-05,
+      "loss": 4.0889,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99655914091209e-05,
+      "loss": 4.1019,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995720546161038e-05,
+      "loss": 4.0867,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994881951409986e-05,
+      "loss": 4.1017,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994043356658934e-05,
+      "loss": 4.0854,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932047619078815e-05,
+      "loss": 4.097,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992367805037203e-05,
+      "loss": 4.0928,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991529210286151e-05,
+      "loss": 4.0991,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9906906155350984e-05,
+      "loss": 4.1041,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9898520207840464e-05,
+      "loss": 4.0917,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890134260329944e-05,
+      "loss": 4.0953,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988174831281943e-05,
+      "loss": 4.0956,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987336236530891e-05,
+      "loss": 4.0836,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986497641779839e-05,
+      "loss": 4.0851,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985659047028787e-05,
+      "loss": 4.0919,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984820452277735e-05,
+      "loss": 4.091,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983981857526683e-05,
+      "loss": 4.0919,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983143262775631e-05,
+      "loss": 4.1033,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982304668024579e-05,
+      "loss": 4.098,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981466073273527e-05,
+      "loss": 4.0998,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980629116402848e-05,
+      "loss": 4.1012,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979790521651796e-05,
+      "loss": 4.0863,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978951926900744e-05,
+      "loss": 4.0959,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978113332149692e-05,
+      "loss": 4.0838,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97727473739864e-05,
+      "loss": 4.0883,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764377805279616e-05,
+      "loss": 4.0899,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9755991857769096e-05,
+      "loss": 4.0876,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747605910258575e-05,
+      "loss": 4.0892,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739219962748055e-05,
+      "loss": 4.0925,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973083401523753e-05,
+      "loss": 4.0901,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972244806772701e-05,
+      "loss": 4.0945,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9714078499020224e-05,
+      "loss": 4.0831,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705692551509704e-05,
+      "loss": 4.097,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969730660399918e-05,
+      "loss": 4.091,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968892065648866e-05,
+      "loss": 4.077,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968053470897814e-05,
+      "loss": 4.0825,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967214876146762e-05,
+      "loss": 4.0778,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96637628139571e-05,
+      "loss": 4.0866,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965539324525031e-05,
+      "loss": 4.0903,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964700729773979e-05,
+      "loss": 4.0868,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963862135022927e-05,
+      "loss": 4.0867,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963025178152248e-05,
+      "loss": 4.0944,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962186583401196e-05,
+      "loss": 4.0906,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961347988650144e-05,
+      "loss": 4.095,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960509393899092e-05,
+      "loss": 4.086,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95967079914804e-05,
+      "loss": 4.0575,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958832204396988e-05,
+      "loss": 4.1051,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957993609645936e-05,
+      "loss": 4.0813,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957155014894884e-05,
+      "loss": 4.0862,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956316420143832e-05,
+      "loss": 4.0825,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95547782539278e-05,
+      "loss": 4.078,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954639230641728e-05,
+      "loss": 4.0777,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95380227377105e-05,
+      "loss": 4.0737,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952963679019998e-05,
+      "loss": 4.0858,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952125084268946e-05,
+      "loss": 4.0778,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951286489517894e-05,
+      "loss": 4.0957,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950449532647215e-05,
+      "loss": 4.0765,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949610937896163e-05,
+      "loss": 4.0705,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487723431451107e-05,
+      "loss": 4.0906,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479337483940587e-05,
+      "loss": 4.0693,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9470967915233796e-05,
+      "loss": 4.0736,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462581967723276e-05,
+      "loss": 4.0886,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454196020212756e-05,
+      "loss": 4.0891,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445810072702235e-05,
+      "loss": 4.0735,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437424125191715e-05,
+      "loss": 4.0749,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429038177681195e-05,
+      "loss": 4.0645,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420652230170675e-05,
+      "loss": 4.0786,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941228266146389e-05,
+      "loss": 4.0979,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403896713953364e-05,
+      "loss": 4.0806,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9395510766442844e-05,
+      "loss": 4.0858,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387124818932324e-05,
+      "loss": 4.0903,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9378738871421804e-05,
+      "loss": 4.0954,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937036930271501e-05,
+      "loss": 4.0745,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936198335520449e-05,
+      "loss": 4.0861,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935359740769397e-05,
+      "loss": 4.0758,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934521146018345e-05,
+      "loss": 4.0779,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933684189147667e-05,
+      "loss": 4.0782,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932845594396615e-05,
+      "loss": 4.0859,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932006999645563e-05,
+      "loss": 4.0796,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931168404894511e-05,
+      "loss": 4.091,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930329810143459e-05,
+      "loss": 4.0754,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929491215392407e-05,
+      "loss": 4.0748,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928652620641355e-05,
+      "loss": 4.0839,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927814025890303e-05,
+      "loss": 4.0797,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926975431139251e-05,
+      "loss": 4.0706,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926138474268572e-05,
+      "loss": 4.08,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925301517397893e-05,
+      "loss": 4.0802,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924462922646841e-05,
+      "loss": 4.0859,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236243278957894e-05,
+      "loss": 4.0772,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9227857331447374e-05,
+      "loss": 4.0676,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219471383936854e-05,
+      "loss": 4.0906,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211085436426334e-05,
+      "loss": 4.085,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202699488915813e-05,
+      "loss": 4.0798,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194313541405293e-05,
+      "loss": 4.0729,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91859439726985e-05,
+      "loss": 4.0711,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917757440399171e-05,
+      "loss": 4.0707,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916918845648119e-05,
+      "loss": 4.0831,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916080250897067e-05,
+      "loss": 4.0879,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915241656146015e-05,
+      "loss": 4.0765,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914403061394963e-05,
+      "loss": 4.0727,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913566104524285e-05,
+      "loss": 4.0652,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912727509773233e-05,
+      "loss": 4.0665,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118905529025536e-05,
+      "loss": 4.0687,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9110519581515016e-05,
+      "loss": 4.0612,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102133634004496e-05,
+      "loss": 4.0813,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093747686493976e-05,
+      "loss": 4.0671,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085361738983456e-05,
+      "loss": 4.0718,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9076975791472936e-05,
+      "loss": 4.0657,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068589843962416e-05,
+      "loss": 4.0774,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060203896451896e-05,
+      "loss": 4.08,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051834327745105e-05,
+      "loss": 4.0682,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043448380234585e-05,
+      "loss": 4.0883,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035062432724065e-05,
+      "loss": 4.0881,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026676485213545e-05,
+      "loss": 4.0749,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018290537703025e-05,
+      "loss": 4.0776,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9009904590192505e-05,
+      "loss": 4.0638,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9001518642681985e-05,
+      "loss": 4.0702,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993132695171465e-05,
+      "loss": 4.0803,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8984746747660945e-05,
+      "loss": 4.0783,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8976377178954154e-05,
+      "loss": 4.0631,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8967991231443634e-05,
+      "loss": 4.0825,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895963804154058e-05,
+      "loss": 4.0818,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895125209403006e-05,
+      "loss": 4.0746,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894286614651954e-05,
+      "loss": 4.0615,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893448019900902e-05,
+      "loss": 4.0722,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89260942514985e-05,
+      "loss": 4.0676,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891770830398798e-05,
+      "loss": 4.0835,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.073,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8900952787770674e-05,
+      "loss": 4.0786,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889256684026015e-05,
+      "loss": 4.0567,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888418089274963e-05,
+      "loss": 4.089,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887579494523911e-05,
+      "loss": 4.0676,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886740899772859e-05,
+      "loss": 4.0823,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885902305021807e-05,
+      "loss": 4.0759,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885063710270755e-05,
+      "loss": 4.0707,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884228391280449e-05,
+      "loss": 4.0754,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883389796529397e-05,
+      "loss": 4.0866,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882551201778345e-05,
+      "loss": 4.0598,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881712607027293e-05,
+      "loss": 4.076,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880874012276241e-05,
+      "loss": 4.0683,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880035417525189e-05,
+      "loss": 4.0677,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879196822774137e-05,
+      "loss": 4.0689,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878358228023085e-05,
+      "loss": 4.0749,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877519633272033e-05,
+      "loss": 4.0746,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876681038520981e-05,
+      "loss": 4.0703,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875842443769929e-05,
+      "loss": 4.0786,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0893096923828125,
+      "eval_runtime": 315.5641,
+      "eval_samples_per_second": 1209.235,
+      "eval_steps_per_second": 37.789,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87500548689925e-05,
+      "loss": 4.0644,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874166892148198e-05,
+      "loss": 4.0648,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873328297397146e-05,
+      "loss": 4.0727,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872489702646094e-05,
+      "loss": 4.0645,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871652745775416e-05,
+      "loss": 4.0744,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870814151024364e-05,
+      "loss": 4.0616,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869975556273312e-05,
+      "loss": 4.0743,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86913696152226e-05,
+      "loss": 4.0616,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683000046515806e-05,
+      "loss": 4.0665,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674614099005286e-05,
+      "loss": 4.0688,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666228151494766e-05,
+      "loss": 4.0696,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657842203984246e-05,
+      "loss": 4.0753,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649472635277455e-05,
+      "loss": 4.078,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641086687766935e-05,
+      "loss": 4.0615,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632700740256415e-05,
+      "loss": 4.0678,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8624314792745895e-05,
+      "loss": 4.0592,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8615928845235375e-05,
+      "loss": 4.0558,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607542897724855e-05,
+      "loss": 4.0656,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599156950214334e-05,
+      "loss": 4.0616,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8590771002703814e-05,
+      "loss": 4.0665,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858240143399703e-05,
+      "loss": 4.0794,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8574015486486503e-05,
+      "loss": 4.0746,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8565629538975983e-05,
+      "loss": 4.0736,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8557243591465463e-05,
+      "loss": 4.0737,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854885764395494e-05,
+      "loss": 4.0584,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854047169644442e-05,
+      "loss": 4.0687,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85320857489339e-05,
+      "loss": 4.0597,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852369980142338e-05,
+      "loss": 4.064,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851533023271659e-05,
+      "loss": 4.0623,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850694428520608e-05,
+      "loss": 4.0629,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849855833769556e-05,
+      "loss": 4.0622,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849017239018504e-05,
+      "loss": 4.0636,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848178644267452e-05,
+      "loss": 4.0657,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847343325277146e-05,
+      "loss": 4.066,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846504730526094e-05,
+      "loss": 4.0605,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845667773655415e-05,
+      "loss": 4.0674,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448291789043626e-05,
+      "loss": 4.07,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439905841533106e-05,
+      "loss": 4.0529,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431519894022586e-05,
+      "loss": 4.058,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423133946512066e-05,
+      "loss": 4.0524,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414747999001546e-05,
+      "loss": 4.0556,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840636205149103e-05,
+      "loss": 4.064,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839797610398051e-05,
+      "loss": 4.0635,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838959015646999e-05,
+      "loss": 4.0633,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838120420895947e-05,
+      "loss": 4.0676,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837283464025268e-05,
+      "loss": 4.0668,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836444869274216e-05,
+      "loss": 4.0667,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835606274523164e-05,
+      "loss": 4.06,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834767679772112e-05,
+      "loss": 4.0363,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83392908502106e-05,
+      "loss": 4.0781,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833090490270008e-05,
+      "loss": 4.0558,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832251895518956e-05,
+      "loss": 4.06,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831413300767904e-05,
+      "loss": 4.0598,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830576343897225e-05,
+      "loss": 4.0523,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829737749146173e-05,
+      "loss": 4.0527,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828899154395121e-05,
+      "loss": 4.0484,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82806055964407e-05,
+      "loss": 4.0596,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272236027733906e-05,
+      "loss": 4.0507,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263850080223386e-05,
+      "loss": 4.0734,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255480511516595e-05,
+      "loss": 4.0522,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247094564006075e-05,
+      "loss": 4.0407,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238708616495555e-05,
+      "loss": 4.0675,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230322668985035e-05,
+      "loss": 4.0491,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8221936721474515e-05,
+      "loss": 4.046,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821355077396399e-05,
+      "loss": 4.063,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820516482645347e-05,
+      "loss": 4.0666,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8196795257746684e-05,
+      "loss": 4.0479,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8188409310236164e-05,
+      "loss": 4.0487,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8180023362725644e-05,
+      "loss": 4.0404,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8171637415215124e-05,
+      "loss": 4.0535,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8163251467704604e-05,
+      "loss": 4.0695,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8154865520194084e-05,
+      "loss": 4.0593,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8146479572683564e-05,
+      "loss": 4.06,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138093625173044e-05,
+      "loss": 4.0607,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812972405646625e-05,
+      "loss": 4.0705,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812135448775946e-05,
+      "loss": 4.0515,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811296854024894e-05,
+      "loss": 4.0652,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810458259273842e-05,
+      "loss": 4.0471,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80961966452279e-05,
+      "loss": 4.0557,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808781069771739e-05,
+      "loss": 4.0539,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807942475020687e-05,
+      "loss": 4.0582,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807103880269635e-05,
+      "loss": 4.058,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806265285518583e-05,
+      "loss": 4.0651,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805426690767531e-05,
+      "loss": 4.0501,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804588096016479e-05,
+      "loss": 4.0508,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037511391458e-05,
+      "loss": 4.0577,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802912544394748e-05,
+      "loss": 4.0596,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802073949643696e-05,
+      "loss": 4.0477,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801235354892644e-05,
+      "loss": 4.052,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800396760141592e-05,
+      "loss": 4.0574,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79955816539054e-05,
+      "loss": 4.0575,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798719570639488e-05,
+      "loss": 4.0551,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797880975888436e-05,
+      "loss": 4.0445,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797044019017757e-05,
+      "loss": 4.0636,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796205424266705e-05,
+      "loss": 4.0613,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953668295156526e-05,
+      "loss": 4.0568,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7945282347646006e-05,
+      "loss": 4.0474,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793692915774295e-05,
+      "loss": 4.0488,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792854321023243e-05,
+      "loss": 4.0448,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792015726272191e-05,
+      "loss": 4.0612,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791177131521139e-05,
+      "loss": 4.0628,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790338536770087e-05,
+      "loss": 4.0503,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789501579899408e-05,
+      "loss": 4.0556,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788662985148356e-05,
+      "loss": 4.0441,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787824390397304e-05,
+      "loss": 4.0454,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786985795646253e-05,
+      "loss": 4.0425,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7861488387755736e-05,
+      "loss": 4.0415,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853102440245216e-05,
+      "loss": 4.0534,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844716492734696e-05,
+      "loss": 4.0445,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836330545224176e-05,
+      "loss": 4.0474,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782794459771365e-05,
+      "loss": 4.0436,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781955865020313e-05,
+      "loss": 4.0493,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781117270269261e-05,
+      "loss": 4.0576,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780278675518209e-05,
+      "loss": 4.0449,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77944171864753e-05,
+      "loss": 4.0626,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778603123896478e-05,
+      "loss": 4.0648,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7777661670257994e-05,
+      "loss": 4.0541,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769275722747474e-05,
+      "loss": 4.0506,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7760889775236954e-05,
+      "loss": 4.0456,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752503827726433e-05,
+      "loss": 4.0504,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744117880215913e-05,
+      "loss": 4.0561,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735731932705393e-05,
+      "loss": 4.0531,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772734598519487e-05,
+      "loss": 4.0417,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771897641648808e-05,
+      "loss": 4.0608,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771059046897756e-05,
+      "loss": 4.0606,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770220452146704e-05,
+      "loss": 4.0445,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769381857395652e-05,
+      "loss": 4.0384,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7685432626446e-05,
+      "loss": 4.0522,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767704667893548e-05,
+      "loss": 4.04,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766866073142496e-05,
+      "loss": 4.0637,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766027478391444e-05,
+      "loss": 4.0474,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765188883640393e-05,
+      "loss": 4.0561,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764351926769714e-05,
+      "loss": 4.0341,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763513332018662e-05,
+      "loss": 4.0586,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76267473726761e-05,
+      "loss": 4.0526,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761837780396931e-05,
+      "loss": 4.0525,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760999185645879e-05,
+      "loss": 4.0532,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760160590894827e-05,
+      "loss": 4.0463,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759321996143775e-05,
+      "loss": 4.0525,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758483401392723e-05,
+      "loss": 4.071,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7576464445220436e-05,
+      "loss": 4.0335,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7568078497709916e-05,
+      "loss": 4.0549,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7559692550199396e-05,
+      "loss": 4.044,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755130660268888e-05,
+      "loss": 4.0423,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754292065517836e-05,
+      "loss": 4.0476,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7534534707667836e-05,
+      "loss": 4.0497,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752616513896105e-05,
+      "loss": 4.0512,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751777919145053e-05,
+      "loss": 4.0489,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750939324394001e-05,
+      "loss": 4.0545,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.073252201080322,
+      "eval_runtime": 319.9549,
+      "eval_samples_per_second": 1192.64,
+      "eval_steps_per_second": 37.271,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501007296429485e-05,
+      "loss": 4.032,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7492621348918965e-05,
+      "loss": 4.0427,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484235401408445e-05,
+      "loss": 4.0479,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475849453897924e-05,
+      "loss": 4.0448,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467463506387404e-05,
+      "loss": 4.0549,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459077558876884e-05,
+      "loss": 4.0415,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450691611366364e-05,
+      "loss": 4.0506,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744230566385585e-05,
+      "loss": 4.0374,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743391971634533e-05,
+      "loss": 4.0459,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742553376883481e-05,
+      "loss": 4.0454,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741714782132429e-05,
+      "loss": 4.0431,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740876187381377e-05,
+      "loss": 4.0562,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740037592630325e-05,
+      "loss": 4.0575,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7391989978792724e-05,
+      "loss": 4.0382,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7383604031282204e-05,
+      "loss": 4.047,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7375218083771684e-05,
+      "loss": 4.0335,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7366832136261164e-05,
+      "loss": 4.0346,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7358446188750644e-05,
+      "loss": 4.0433,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7350060241240124e-05,
+      "loss": 4.0373,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7341674293729604e-05,
+      "loss": 4.0447,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7333288346219084e-05,
+      "loss": 4.0574,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73249187775123e-05,
+      "loss": 4.0499,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731653283000178e-05,
+      "loss": 4.0531,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730816326129499e-05,
+      "loss": 4.0494,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729977731378447e-05,
+      "loss": 4.04,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729139136627395e-05,
+      "loss": 4.0435,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728300541876343e-05,
+      "loss": 4.0372,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727461947125291e-05,
+      "loss": 4.0452,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726623352374239e-05,
+      "loss": 4.0375,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725784757623187e-05,
+      "loss": 4.0413,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724946162872135e-05,
+      "loss": 4.0404,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724109206001456e-05,
+      "loss": 4.0433,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723270611250404e-05,
+      "loss": 4.0465,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722433654379725e-05,
+      "loss": 4.0404,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721595059628673e-05,
+      "loss": 4.0399,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720758102757994e-05,
+      "loss": 4.0435,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719919508006942e-05,
+      "loss": 4.0491,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71908091325589e-05,
+      "loss": 4.0321,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718242318504838e-05,
+      "loss": 4.033,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717403723753786e-05,
+      "loss": 4.033,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716565129002734e-05,
+      "loss": 4.0326,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715726534251682e-05,
+      "loss": 4.0394,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71488793950063e-05,
+      "loss": 4.0429,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714050982629951e-05,
+      "loss": 4.0413,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713212387878899e-05,
+      "loss": 4.0447,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712373793127847e-05,
+      "loss": 4.0446,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711535198376795e-05,
+      "loss": 4.0453,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710696603625744e-05,
+      "loss": 4.0409,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709858008874691e-05,
+      "loss": 4.0128,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709019414123639e-05,
+      "loss": 4.0544,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708180819372587e-05,
+      "loss": 4.0317,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707342224621535e-05,
+      "loss": 4.0377,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706505267750856e-05,
+      "loss": 4.0392,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7056683108801776e-05,
+      "loss": 4.0258,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048297161291256e-05,
+      "loss": 4.0361,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703991121378073e-05,
+      "loss": 4.027,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703152526627021e-05,
+      "loss": 4.0418,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702313931875969e-05,
+      "loss": 4.0276,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014753371249176e-05,
+      "loss": 4.0501,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006367423738655e-05,
+      "loss": 4.0324,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69980142338356e-05,
+      "loss": 4.0183,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698962828632508e-05,
+      "loss": 4.0493,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698124233881456e-05,
+      "loss": 4.0254,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972856391304034e-05,
+      "loss": 4.0233,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6964470443793514e-05,
+      "loss": 4.0423,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6956084496282993e-05,
+      "loss": 4.0463,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6947698548772473e-05,
+      "loss": 4.0259,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939312601261953e-05,
+      "loss": 4.0282,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693092665375143e-05,
+      "loss": 4.02,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692254070624091e-05,
+      "loss": 4.0287,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691415475873039e-05,
+      "loss": 4.0448,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690576881121987e-05,
+      "loss": 4.0391,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689738286370936e-05,
+      "loss": 4.039,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688901329500257e-05,
+      "loss": 4.0393,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688062734749205e-05,
+      "loss": 4.0486,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687225777878526e-05,
+      "loss": 4.0355,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686387183127474e-05,
+      "loss": 4.041,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685548588376422e-05,
+      "loss": 4.025,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68470999362537e-05,
+      "loss": 4.0325,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683871398874318e-05,
+      "loss": 4.0327,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683032804123266e-05,
+      "loss": 4.0359,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682194209372214e-05,
+      "loss": 4.0406,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681355614621162e-05,
+      "loss": 4.0451,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680518657750483e-05,
+      "loss": 4.032,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679680062999431e-05,
+      "loss": 4.0274,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6788414682483794e-05,
+      "loss": 4.035,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6780028734973274e-05,
+      "loss": 4.0391,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677164278746275e-05,
+      "loss": 4.0266,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676327321875596e-05,
+      "loss": 4.0324,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675490365004917e-05,
+      "loss": 4.041,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674651770253865e-05,
+      "loss": 4.0337,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673813175502813e-05,
+      "loss": 4.034,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672974580751761e-05,
+      "loss": 4.0244,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672135986000709e-05,
+      "loss": 4.0392,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6712973912496565e-05,
+      "loss": 4.0408,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704587964986045e-05,
+      "loss": 4.035,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669620201747553e-05,
+      "loss": 4.0293,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668781606996501e-05,
+      "loss": 4.0267,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667944650125822e-05,
+      "loss": 4.0259,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66710605537477e-05,
+      "loss": 4.0381,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662690985040916e-05,
+      "loss": 4.043,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654305037530396e-05,
+      "loss": 4.0268,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664591909001987e-05,
+      "loss": 4.0311,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6637549521313085e-05,
+      "loss": 4.0279,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629163573802565e-05,
+      "loss": 4.0195,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662077762629204e-05,
+      "loss": 4.0233,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661239167878152e-05,
+      "loss": 4.0209,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604005731271e-05,
+      "loss": 4.0264,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595619783760485e-05,
+      "loss": 4.027,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587233836249965e-05,
+      "loss": 4.0223,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6578847888739445e-05,
+      "loss": 4.0268,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6570461941228925e-05,
+      "loss": 4.0262,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6562092372522134e-05,
+      "loss": 4.0387,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6553706425011614e-05,
+      "loss": 4.0219,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6545320477501094e-05,
+      "loss": 4.0447,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6536934529990574e-05,
+      "loss": 4.0421,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652856496128378e-05,
+      "loss": 4.037,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652017901377326e-05,
+      "loss": 4.0293,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651179306626274e-05,
+      "loss": 4.0266,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650340711875222e-05,
+      "loss": 4.0295,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64950211712417e-05,
+      "loss": 4.0384,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648663522373118e-05,
+      "loss": 4.0331,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647824927622067e-05,
+      "loss": 4.0234,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646986332871015e-05,
+      "loss": 4.0344,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646151013880709e-05,
+      "loss": 4.0411,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645312419129657e-05,
+      "loss": 4.0242,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644473824378605e-05,
+      "loss": 4.0201,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643635229627553e-05,
+      "loss": 4.0307,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642796634876501e-05,
+      "loss": 4.0192,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641958040125449e-05,
+      "loss": 4.0448,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641119445374397e-05,
+      "loss": 4.0261,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640280850623345e-05,
+      "loss": 4.0366,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6394438937526656e-05,
+      "loss": 4.0147,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386052990016136e-05,
+      "loss": 4.0367,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637768342130935e-05,
+      "loss": 4.0311,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636929747379883e-05,
+      "loss": 4.0327,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636091152628831e-05,
+      "loss": 4.0307,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635252557877779e-05,
+      "loss": 4.0255,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634413963126727e-05,
+      "loss": 4.0317,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633575368375675e-05,
+      "loss": 4.0501,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6327367736246225e-05,
+      "loss": 4.0161,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631899816753944e-05,
+      "loss": 4.0304,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631061222002892e-05,
+      "loss": 4.0297,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63022262725184e-05,
+      "loss": 4.0225,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293840325007874e-05,
+      "loss": 4.0269,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628547075630109e-05,
+      "loss": 4.0281,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627708480879058e-05,
+      "loss": 4.0303,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6268715240083786e-05,
+      "loss": 4.0294,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6260329292573266e-05,
+      "loss": 4.0336,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.060416221618652,
+      "eval_runtime": 340.3783,
+      "eval_samples_per_second": 1121.079,
+      "eval_steps_per_second": 35.035,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251943345062746e-05,
+      "loss": 4.0207,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243557397552226e-05,
+      "loss": 4.0217,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62351714500417e-05,
+      "loss": 4.0311,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622678550253118e-05,
+      "loss": 4.0241,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621839955502066e-05,
+      "loss": 4.032,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621001360751014e-05,
+      "loss": 4.0228,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620162765999962e-05,
+      "loss": 4.0288,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619325809129283e-05,
+      "loss": 4.0232,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6184872143782315e-05,
+      "loss": 4.0261,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176486196271795e-05,
+      "loss": 4.0202,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168116627565004e-05,
+      "loss": 4.0245,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6159730680054484e-05,
+      "loss": 4.0393,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6151344732543964e-05,
+      "loss": 4.0329,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6142958785033444e-05,
+      "loss": 4.019,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6134572837522923e-05,
+      "loss": 4.0274,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6126186890012403e-05,
+      "loss": 4.0141,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611780094250188e-05,
+      "loss": 4.0174,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610943137379509e-05,
+      "loss": 4.029,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610104542628457e-05,
+      "loss": 4.0143,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609265947877405e-05,
+      "loss": 4.0244,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608427353126353e-05,
+      "loss": 4.0333,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607588758375301e-05,
+      "loss": 4.0322,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606750163624249e-05,
+      "loss": 4.0331,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605913206753571e-05,
+      "loss": 4.0309,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605074612002519e-05,
+      "loss": 4.0196,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604236017251467e-05,
+      "loss": 4.0242,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603397422500415e-05,
+      "loss": 4.0168,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602558827749363e-05,
+      "loss": 4.0247,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601720232998311e-05,
+      "loss": 4.0197,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600881638247259e-05,
+      "loss": 4.0232,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600043043496206e-05,
+      "loss": 4.0228,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599206086625528e-05,
+      "loss": 4.0203,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598367491874476e-05,
+      "loss": 4.0239,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597528897123424e-05,
+      "loss": 4.0255,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596690302372372e-05,
+      "loss": 4.0197,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595853345501693e-05,
+      "loss": 4.0262,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595014750750641e-05,
+      "loss": 4.0317,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594176155999589e-05,
+      "loss": 4.0121,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5933375612485366e-05,
+      "loss": 4.0141,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5924989664974846e-05,
+      "loss": 4.0147,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5916603717464326e-05,
+      "loss": 4.0113,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908217769953806e-05,
+      "loss": 4.0226,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899831822443286e-05,
+      "loss": 4.0203,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891462253736495e-05,
+      "loss": 4.0275,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5883076306225975e-05,
+      "loss": 4.0225,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874690358715455e-05,
+      "loss": 4.0261,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5866304411204935e-05,
+      "loss": 4.0247,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5857918463694414e-05,
+      "loss": 4.0225,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58495325161839e-05,
+      "loss": 3.997,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584114656867338e-05,
+      "loss": 4.0326,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583276062116286e-05,
+      "loss": 4.0132,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582439105245607e-05,
+      "loss": 4.0188,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581600510494555e-05,
+      "loss": 4.0153,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580761915743503e-05,
+      "loss": 4.0087,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579923320992451e-05,
+      "loss": 4.0174,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579086364121772e-05,
+      "loss": 4.0107,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57824776937072e-05,
+      "loss": 4.0219,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577410812500041e-05,
+      "loss": 4.0063,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576572217748989e-05,
+      "loss": 4.0327,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575733622997937e-05,
+      "loss": 4.0197,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5748950282468855e-05,
+      "loss": 3.9936,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5740564334958335e-05,
+      "loss": 4.0296,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5732178387447815e-05,
+      "loss": 4.0063,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5723808818741024e-05,
+      "loss": 4.0082,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5715422871230504e-05,
+      "loss": 4.0203,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5707036923719984e-05,
+      "loss": 4.0267,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5698650976209464e-05,
+      "loss": 4.0073,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690265028698944e-05,
+      "loss": 4.0161,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5681879081188424e-05,
+      "loss": 3.9996,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56734931336779e-05,
+      "loss": 4.0086,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566510718616738e-05,
+      "loss": 4.0288,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565673761746059e-05,
+      "loss": 4.0198,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564835166995007e-05,
+      "loss": 4.0225,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563996572243955e-05,
+      "loss": 4.0208,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563157977492903e-05,
+      "loss": 4.0306,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562319382741851e-05,
+      "loss": 4.0175,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561482425871172e-05,
+      "loss": 4.0236,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56064383112012e-05,
+      "loss": 4.0068,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559805236369068e-05,
+      "loss": 4.0132,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558966641618016e-05,
+      "loss": 4.0134,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558128046866964e-05,
+      "loss": 4.0156,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557289452115912e-05,
+      "loss": 4.0178,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55645085736486e-05,
+      "loss": 4.0263,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555613900494181e-05,
+      "loss": 4.0193,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547769436235026e-05,
+      "loss": 4.005,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5539383488724506e-05,
+      "loss": 4.0187,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5530997541213986e-05,
+      "loss": 4.0208,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522611593703466e-05,
+      "loss": 4.0066,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5514225646192946e-05,
+      "loss": 4.0147,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5505856077486155e-05,
+      "loss": 4.0222,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5497470129975635e-05,
+      "loss": 4.0133,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489084182465115e-05,
+      "loss": 4.0194,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5480698234954595e-05,
+      "loss": 4.005,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5472312287444075e-05,
+      "loss": 4.0232,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5463926339933555e-05,
+      "loss": 4.0209,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5455540392423035e-05,
+      "loss": 4.0233,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5447154444912515e-05,
+      "loss": 4.0037,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438784876205724e-05,
+      "loss": 4.0148,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543039892869521e-05,
+      "loss": 4.0076,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542201298118469e-05,
+      "loss": 4.0182,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541362703367417e-05,
+      "loss": 4.0282,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540524108616365e-05,
+      "loss": 4.0077,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539687151745686e-05,
+      "loss": 4.0159,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538848556994634e-05,
+      "loss": 4.0062,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538009962243582e-05,
+      "loss": 4.0045,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537173005372903e-05,
+      "loss": 4.0062,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536334410621851e-05,
+      "loss": 4.0043,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535495815870799e-05,
+      "loss": 4.0104,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534657221119747e-05,
+      "loss": 4.0091,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533818626368695e-05,
+      "loss": 4.0046,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5329816694980164e-05,
+      "loss": 4.0025,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321430747469644e-05,
+      "loss": 4.0073,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313061178762853e-05,
+      "loss": 4.0237,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304675231252333e-05,
+      "loss": 4.0094,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529628928374181e-05,
+      "loss": 4.0229,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528790333623129e-05,
+      "loss": 4.0206,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527951738872077e-05,
+      "loss": 4.0248,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527113144121025e-05,
+      "loss": 4.0107,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526274549369973e-05,
+      "loss": 4.0083,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254359546189206e-05,
+      "loss": 4.0121,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5245973598678686e-05,
+      "loss": 4.0182,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237587651168166e-05,
+      "loss": 4.0171,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522921808246138e-05,
+      "loss": 4.0091,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522083213495086e-05,
+      "loss": 4.0102,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521244618744034e-05,
+      "loss": 4.0235,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520406023992982e-05,
+      "loss": 4.0127,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51956742924193e-05,
+      "loss": 4.0011,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518728834490878e-05,
+      "loss": 4.0116,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517890239739826e-05,
+      "loss": 4.0028,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517051644988774e-05,
+      "loss": 4.0246,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516214688118095e-05,
+      "loss": 4.0085,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515376093367043e-05,
+      "loss": 4.0189,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514537498615991e-05,
+      "loss": 3.9995,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513698903864939e-05,
+      "loss": 4.018,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51286194699426e-05,
+      "loss": 4.0134,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512023352243209e-05,
+      "loss": 4.0139,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111847574921567e-05,
+      "loss": 4.015,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103461627411047e-05,
+      "loss": 4.0073,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095108437507985e-05,
+      "loss": 4.0176,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086722489997465e-05,
+      "loss": 4.0325,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5078336542486945e-05,
+      "loss": 4.0003,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5069950594976425e-05,
+      "loss": 4.0109,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5061564647465905e-05,
+      "loss": 4.0133,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053195078759114e-05,
+      "loss": 4.0071,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044809131248594e-05,
+      "loss": 4.0069,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036423183738074e-05,
+      "loss": 4.0143,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028037236227554e-05,
+      "loss": 4.0114,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501965128871704e-05,
+      "loss": 4.011,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501126534120652e-05,
+      "loss": 4.0179,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.049440383911133,
+      "eval_runtime": 340.5482,
+      "eval_samples_per_second": 1120.52,
+      "eval_steps_per_second": 35.017,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002879393696e-05,
+      "loss": 4.0014,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499449344618548e-05,
+      "loss": 4.0011,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498610749867496e-05,
+      "loss": 4.0134,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497772155116444e-05,
+      "loss": 4.0097,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496933560365392e-05,
+      "loss": 4.012,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496094965614339e-05,
+      "loss": 4.0045,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495256370863287e-05,
+      "loss": 4.0159,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494417776112235e-05,
+      "loss": 4.0038,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493579181361183e-05,
+      "loss": 4.0084,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492740586610131e-05,
+      "loss": 4.0039,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491901991859079e-05,
+      "loss": 4.0091,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491063397108027e-05,
+      "loss": 4.019,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490226440237349e-05,
+      "loss": 4.02,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489387845486297e-05,
+      "loss": 3.9995,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488549250735245e-05,
+      "loss": 4.0119,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487710655984193e-05,
+      "loss": 3.995,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486872061233141e-05,
+      "loss": 4.0035,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486035104362462e-05,
+      "loss": 4.0057,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48519650961141e-05,
+      "loss": 3.9982,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484357914860358e-05,
+      "loss": 4.0078,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483519320109306e-05,
+      "loss": 4.0158,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482680725358254e-05,
+      "loss": 4.0099,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481842130607202e-05,
+      "loss": 4.0198,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48100353585615e-05,
+      "loss": 4.0172,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480164941105098e-05,
+      "loss": 4.0,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479326346354046e-05,
+      "loss": 4.0103,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478487751602994e-05,
+      "loss": 3.999,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477650794732315e-05,
+      "loss": 4.0061,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476812199981263e-05,
+      "loss": 4.0026,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759736052302106e-05,
+      "loss": 4.0042,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751350104791586e-05,
+      "loss": 4.01,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4742964157281066e-05,
+      "loss": 4.002,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734578209770546e-05,
+      "loss": 4.0061,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726192262260026e-05,
+      "loss": 4.0122,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717806314749506e-05,
+      "loss": 4.0025,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709436746042715e-05,
+      "loss": 4.0074,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4701050798532195e-05,
+      "loss": 4.0157,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4692664851021675e-05,
+      "loss": 3.9908,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4684278903511155e-05,
+      "loss": 4.0019,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467592571360811e-05,
+      "loss": 3.9989,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466753976609758e-05,
+      "loss": 3.9934,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465915381858706e-05,
+      "loss": 4.0035,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465076787107654e-05,
+      "loss": 4.0037,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464238192356602e-05,
+      "loss": 4.0089,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46339959760555e-05,
+      "loss": 4.0066,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462561002854498e-05,
+      "loss": 4.0087,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461724045983819e-05,
+      "loss": 4.0114,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460885451232767e-05,
+      "loss": 4.0044,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460046856481715e-05,
+      "loss": 3.9843,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459208261730663e-05,
+      "loss": 4.0107,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458369666979611e-05,
+      "loss": 3.9979,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575310722285595e-05,
+      "loss": 4.0004,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566924774775075e-05,
+      "loss": 4.0003,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558538827264555e-05,
+      "loss": 3.9917,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550169258557764e-05,
+      "loss": 4.0003,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.9913,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533397363536724e-05,
+      "loss": 4.0066,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525011416026204e-05,
+      "loss": 3.987,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516641847319413e-05,
+      "loss": 4.0168,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508255899808893e-05,
+      "loss": 4.0032,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449986995229837e-05,
+      "loss": 3.9806,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449148400478785e-05,
+      "loss": 4.0122,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448309805727733e-05,
+      "loss": 3.9876,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447471210976681e-05,
+      "loss": 3.9944,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 3.9998,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 4.0117,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444957064603899e-05,
+      "loss": 3.9921,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444118469852847e-05,
+      "loss": 3.9987,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443279875101794e-05,
+      "loss": 3.9778,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442441280350742e-05,
+      "loss": 3.9983,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44160268559969e-05,
+      "loss": 4.0097,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440764090848638e-05,
+      "loss": 4.0072,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439927133977959e-05,
+      "loss": 4.0028,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439088539226907e-05,
+      "loss": 4.0045,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438249944475855e-05,
+      "loss": 4.0164,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437412987605177e-05,
+      "loss": 3.9999,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436576030734498e-05,
+      "loss": 4.0033,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435737435983446e-05,
+      "loss": 3.9938,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434898841232394e-05,
+      "loss": 3.9998,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340602464813416e-05,
+      "loss": 3.9949,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433223289610663e-05,
+      "loss": 3.9969,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432384694859611e-05,
+      "loss": 4.0064,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431546100108559e-05,
+      "loss": 4.0058,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307075053575065e-05,
+      "loss": 4.0099,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298689106064545e-05,
+      "loss": 3.9867,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4290303158554025e-05,
+      "loss": 4.0018,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281933589847234e-05,
+      "loss": 4.0037,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427354764233672e-05,
+      "loss": 3.9924,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42651616948262e-05,
+      "loss": 3.9964,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425677574731568e-05,
+      "loss": 4.0043,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424838979980516e-05,
+      "loss": 4.0016,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424000385229464e-05,
+      "loss": 3.998,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423161790478412e-05,
+      "loss": 3.9934,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42232319572736e-05,
+      "loss": 4.0034,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421484600976308e-05,
+      "loss": 4.0083,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420647644105629e-05,
+      "loss": 4.003,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419809049354577e-05,
+      "loss": 3.9897,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418970454603525e-05,
+      "loss": 3.9971,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418131859852473e-05,
+      "loss": 3.9921,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417293265101421e-05,
+      "loss": 4.0033,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416454670350369e-05,
+      "loss": 4.0125,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415616075599317e-05,
+      "loss": 3.9908,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147791187286385e-05,
+      "loss": 3.9981,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139405239775865e-05,
+      "loss": 3.9932,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131019292265345e-05,
+      "loss": 3.9912,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122633344754825e-05,
+      "loss": 3.9855,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114263776048034e-05,
+      "loss": 3.9929,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105877828537514e-05,
+      "loss": 3.9923,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097491881026994e-05,
+      "loss": 3.9937,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4089105933516474e-05,
+      "loss": 3.9881,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408071998600595e-05,
+      "loss": 3.9896,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407233403849543e-05,
+      "loss": 3.9887,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406394809098491e-05,
+      "loss": 4.0066,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405556214347439e-05,
+      "loss": 3.9908,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047176195963874e-05,
+      "loss": 4.0077,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403880662725708e-05,
+      "loss": 4.0022,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403042067974656e-05,
+      "loss": 4.0103,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9976,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401366516352925e-05,
+      "loss": 3.9911,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400527921601873e-05,
+      "loss": 3.9959,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399689326850821e-05,
+      "loss": 4.0011,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398850732099769e-05,
+      "loss": 3.9972,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398012137348717e-05,
+      "loss": 3.9981,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397173542597665e-05,
+      "loss": 3.9928,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396334947846613e-05,
+      "loss": 4.0085,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395497990975934e-05,
+      "loss": 3.9919,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394659396224882e-05,
+      "loss": 3.9901,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393820801473831e-05,
+      "loss": 3.9953,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392982206722779e-05,
+      "loss": 3.9921,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921452498520996e-05,
+      "loss": 4.009,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913066551010476e-05,
+      "loss": 3.9921,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904680603499956e-05,
+      "loss": 4.0006,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896294655989436e-05,
+      "loss": 3.988,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3887925087282645e-05,
+      "loss": 4.0018,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879539139772125e-05,
+      "loss": 3.9997,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871153192261605e-05,
+      "loss": 4.0,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862767244751085e-05,
+      "loss": 3.9921,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854381297240565e-05,
+      "loss": 3.9995,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3846011728533774e-05,
+      "loss": 4.002,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383762578102326e-05,
+      "loss": 4.0135,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382923983351274e-05,
+      "loss": 3.9888,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382085388600222e-05,
+      "loss": 3.9933,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38124679384917e-05,
+      "loss": 3.9962,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380408199098118e-05,
+      "loss": 3.9915,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379569604347066e-05,
+      "loss": 3.9902,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378731009596014e-05,
+      "loss": 4.0006,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377894052725335e-05,
+      "loss": 3.9921,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 4.0028,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376218501103604e-05,
+      "loss": 3.9965,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.039336204528809,
+      "eval_runtime": 304.4595,
+      "eval_samples_per_second": 1253.339,
+      "eval_steps_per_second": 39.168,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375379906352552e-05,
+      "loss": 3.9945,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3745413116015e-05,
+      "loss": 3.9823,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3737043547308215e-05,
+      "loss": 3.9963,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728657599797694e-05,
+      "loss": 3.9978,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720271652287174e-05,
+      "loss": 3.9951,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3711885704776654e-05,
+      "loss": 3.9923,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703499757266134e-05,
+      "loss": 3.9958,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695113809755614e-05,
+      "loss": 3.9922,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368672786224509e-05,
+      "loss": 3.9889,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36783582935383e-05,
+      "loss": 3.9957,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366997234602778e-05,
+      "loss": 3.9913,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3661586398517256e-05,
+      "loss": 4.0006,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365321682981047e-05,
+      "loss": 4.0102,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364483088229995e-05,
+      "loss": 3.9826,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363644493478943e-05,
+      "loss": 3.9905,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362805898727891e-05,
+      "loss": 3.9834,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361967303976839e-05,
+      "loss": 3.9891,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361128709225787e-05,
+      "loss": 3.9919,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360290114474735e-05,
+      "loss": 3.9832,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359451519723683e-05,
+      "loss": 3.9885,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358612924972631e-05,
+      "loss": 4.0058,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357774330221579e-05,
+      "loss": 3.9926,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356935735470527e-05,
+      "loss": 4.0018,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356097140719475e-05,
+      "loss": 4.0004,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355261821729169e-05,
+      "loss": 3.9864,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354423226978117e-05,
+      "loss": 3.9967,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353584632227065e-05,
+      "loss": 3.9824,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352746037476014e-05,
+      "loss": 3.9906,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351907442724962e-05,
+      "loss": 3.9892,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35106884797391e-05,
+      "loss": 3.9867,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350230253222858e-05,
+      "loss": 3.9966,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3493916584718057e-05,
+      "loss": 3.9864,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485547016011266e-05,
+      "loss": 3.994,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477177447304475e-05,
+      "loss": 3.9894,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468791499793955e-05,
+      "loss": 3.9902,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460421931087164e-05,
+      "loss": 3.9958,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3452035983576644e-05,
+      "loss": 4.0017,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443650036066124e-05,
+      "loss": 3.9715,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435264088555604e-05,
+      "loss": 3.9898,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342687814104509e-05,
+      "loss": 3.9865,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341849219353457e-05,
+      "loss": 3.9764,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341010624602405e-05,
+      "loss": 3.9874,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340172029851353e-05,
+      "loss": 3.994,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339335072980674e-05,
+      "loss": 3.9931,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338496478229622e-05,
+      "loss": 3.9936,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33765788347857e-05,
+      "loss": 3.9938,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336819288727518e-05,
+      "loss": 3.9932,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335980693976466e-05,
+      "loss": 3.9897,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335142099225414e-05,
+      "loss": 3.9706,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334305142354735e-05,
+      "loss": 3.9938,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333466547603683e-05,
+      "loss": 3.9862,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332627952852631e-05,
+      "loss": 3.9888,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331789358101579e-05,
+      "loss": 3.9824,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309524012309004e-05,
+      "loss": 3.9789,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301138064798484e-05,
+      "loss": 3.9831,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292752117287964e-05,
+      "loss": 3.975,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284366169777444e-05,
+      "loss": 3.9935,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327598022226692e-05,
+      "loss": 3.9757,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326761065356013e-05,
+      "loss": 3.9955,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325922470604961e-05,
+      "loss": 3.9915,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325083875853909e-05,
+      "loss": 3.9676,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3242452811028566e-05,
+      "loss": 3.993,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3234066863518046e-05,
+      "loss": 3.9757,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322569729481126e-05,
+      "loss": 3.9798,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321731134730074e-05,
+      "loss": 3.9814,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320892539979022e-05,
+      "loss": 3.9965,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32005394522797e-05,
+      "loss": 3.9813,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319215350476918e-05,
+      "loss": 3.9816,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318376755725866e-05,
+      "loss": 3.9647,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317538160974814e-05,
+      "loss": 3.98,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316699566223762e-05,
+      "loss": 3.9962,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315862609353083e-05,
+      "loss": 3.995,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315024014602031e-05,
+      "loss": 3.9842,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314185419850979e-05,
+      "loss": 3.9936,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313346825099927e-05,
+      "loss": 3.9979,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312508230348875e-05,
+      "loss": 3.9835,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311671273478196e-05,
+      "loss": 3.986,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108326787271446e-05,
+      "loss": 3.9852,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3099940839760926e-05,
+      "loss": 3.9827,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3091554892250406e-05,
+      "loss": 3.9785,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3083168944739886e-05,
+      "loss": 3.9841,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3074782997229366e-05,
+      "loss": 3.9911,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3066397049718846e-05,
+      "loss": 3.9928,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058011102208326e-05,
+      "loss": 3.9933,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049641533501535e-05,
+      "loss": 3.9731,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3041271964794744e-05,
+      "loss": 3.9886,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3032886017284224e-05,
+      "loss": 3.9878,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3024500069773704e-05,
+      "loss": 3.9791,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3016114122263184e-05,
+      "loss": 3.9803,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30077445535564e-05,
+      "loss": 3.9874,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299935860604588e-05,
+      "loss": 3.9852,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299097265853536e-05,
+      "loss": 3.986,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298258671102484e-05,
+      "loss": 3.9784,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297420076351432e-05,
+      "loss": 3.9908,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29658148160038e-05,
+      "loss": 3.9913,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295742886849328e-05,
+      "loss": 3.9891,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294904292098275e-05,
+      "loss": 3.9711,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294065697347223e-05,
+      "loss": 3.9856,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293227102596171e-05,
+      "loss": 3.9791,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292388507845119e-05,
+      "loss": 3.9849,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29155155097444e-05,
+      "loss": 4.0008,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290712956223388e-05,
+      "loss": 3.9762,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289874361472337e-05,
+      "loss": 3.9824,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289035766721285e-05,
+      "loss": 3.9754,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288197171970233e-05,
+      "loss": 3.983,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287360215099554e-05,
+      "loss": 3.9698,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286521620348502e-05,
+      "loss": 3.9765,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28568302559745e-05,
+      "loss": 3.9788,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284844430846398e-05,
+      "loss": 3.9803,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284005836095346e-05,
+      "loss": 3.9771,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283167241344294e-05,
+      "loss": 3.9754,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823302844736146e-05,
+      "loss": 3.9726,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2814916897225626e-05,
+      "loss": 3.9908,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2806530949715106e-05,
+      "loss": 3.9789,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2798145002204586e-05,
+      "loss": 3.9897,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2789759054694066e-05,
+      "loss": 3.9903,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2781373107183546e-05,
+      "loss": 3.999,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277300353847676e-05,
+      "loss": 3.9827,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276461759096624e-05,
+      "loss": 3.9772,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275623164345572e-05,
+      "loss": 3.9833,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27478456959452e-05,
+      "loss": 3.983,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273945974843468e-05,
+      "loss": 3.9823,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273107380092416e-05,
+      "loss": 3.9864,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272268785341364e-05,
+      "loss": 3.9801,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271431828470685e-05,
+      "loss": 3.9903,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270593233719633e-05,
+      "loss": 3.9808,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269754638968581e-05,
+      "loss": 3.9755,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268916044217529e-05,
+      "loss": 3.9804,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268077449466477e-05,
+      "loss": 3.9781,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267238854715425e-05,
+      "loss": 3.9883,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266400259964373e-05,
+      "loss": 3.9835,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265563303093694e-05,
+      "loss": 3.984,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264724708342642e-05,
+      "loss": 3.9769,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26388611359159e-05,
+      "loss": 3.9855,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263047518840538e-05,
+      "loss": 3.9891,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262208924089486e-05,
+      "loss": 3.9826,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261370329338434e-05,
+      "loss": 3.98,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260533372467755e-05,
+      "loss": 3.9835,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259694777716703e-05,
+      "loss": 3.9873,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258856182965651e-05,
+      "loss": 4.0024,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258017588214599e-05,
+      "loss": 3.9768,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2571806313439204e-05,
+      "loss": 3.978,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563420365928684e-05,
+      "loss": 3.9797,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2555034418418164e-05,
+      "loss": 3.9803,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2546648470907644e-05,
+      "loss": 3.9717,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2538262523397124e-05,
+      "loss": 3.9884,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252989295469033e-05,
+      "loss": 3.979,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252150700717981e-05,
+      "loss": 3.9881,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251312105966929e-05,
+      "loss": 3.979,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.031044006347656,
+      "eval_runtime": 290.529,
+      "eval_samples_per_second": 1313.435,
+      "eval_steps_per_second": 41.046,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250473511215877e-05,
+      "loss": 3.9597,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249634916464825e-05,
+      "loss": 3.9736,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248796321713773e-05,
+      "loss": 3.9808,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247957726962721e-05,
+      "loss": 3.984,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247119132211669e-05,
+      "loss": 3.9821,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246280537460617e-05,
+      "loss": 3.9779,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245441942709565e-05,
+      "loss": 3.979,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244603347958513e-05,
+      "loss": 3.9775,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243764753207461e-05,
+      "loss": 3.9762,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242926158456409e-05,
+      "loss": 3.9809,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242087563705357e-05,
+      "loss": 3.978,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241248968954305e-05,
+      "loss": 3.9905,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240412012083626e-05,
+      "loss": 3.9934,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239573417332574e-05,
+      "loss": 3.9676,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238734822581522e-05,
+      "loss": 3.9796,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23789622783047e-05,
+      "loss": 3.9677,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237057633079418e-05,
+      "loss": 3.9743,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236219038328366e-05,
+      "loss": 3.9763,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235380443577314e-05,
+      "loss": 3.9734,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234541848826262e-05,
+      "loss": 3.9698,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23370325407521e-05,
+      "loss": 3.9955,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232864659324159e-05,
+      "loss": 3.9799,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232026064573107e-05,
+      "loss": 3.9881,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231187469822054e-05,
+      "loss": 3.9865,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230350512951376e-05,
+      "loss": 3.9799,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2295135560806966e-05,
+      "loss": 3.9791,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2286749613296446e-05,
+      "loss": 3.9711,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2278363665785926e-05,
+      "loss": 3.9759,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2269977718275406e-05,
+      "loss": 3.9747,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2261591770764886e-05,
+      "loss": 3.9742,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2253205823254366e-05,
+      "loss": 3.9808,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224481987574384e-05,
+      "loss": 3.9704,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2236433928233326e-05,
+      "loss": 3.9821,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222806435952654e-05,
+      "loss": 3.9755,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2219678412016015e-05,
+      "loss": 3.9753,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221130884330923e-05,
+      "loss": 3.9876,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220293927460244e-05,
+      "loss": 3.9847,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219455332709192e-05,
+      "loss": 3.9588,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21861673795814e-05,
+      "loss": 3.9756,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217778143207088e-05,
+      "loss": 3.9711,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216939548456036e-05,
+      "loss": 3.961,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216100953704984e-05,
+      "loss": 3.9787,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215262358953931e-05,
+      "loss": 3.9843,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214423764202879e-05,
+      "loss": 3.9744,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213585169451828e-05,
+      "loss": 3.9802,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212746574700776e-05,
+      "loss": 3.9805,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211907979949724e-05,
+      "loss": 3.9777,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211071023079045e-05,
+      "loss": 3.9779,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210232428327993e-05,
+      "loss": 3.9614,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209393833576941e-05,
+      "loss": 3.9779,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208555238825889e-05,
+      "loss": 3.9749,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207716644074837e-05,
+      "loss": 3.9746,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206878049323785e-05,
+      "loss": 3.9683,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206039454572733e-05,
+      "loss": 3.9682,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205200859821681e-05,
+      "loss": 3.9636,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2043655408313746e-05,
+      "loss": 3.964,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203526946080323e-05,
+      "loss": 3.9836,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202688351329271e-05,
+      "loss": 3.9601,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201849756578219e-05,
+      "loss": 3.9813,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201011161827167e-05,
+      "loss": 3.9762,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200172567076115e-05,
+      "loss": 3.9603,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199333972325063e-05,
+      "loss": 3.9735,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198495377574011e-05,
+      "loss": 3.9654,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197658420703332e-05,
+      "loss": 3.9695,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196821463832653e-05,
+      "loss": 3.9641,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195982869081601e-05,
+      "loss": 3.9846,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195144274330549e-05,
+      "loss": 3.97,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194305679579497e-05,
+      "loss": 3.9628,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193467084828445e-05,
+      "loss": 3.9554,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192628490077393e-05,
+      "loss": 3.9635,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191789895326342e-05,
+      "loss": 3.9806,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190952938455663e-05,
+      "loss": 3.9869,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190114343704611e-05,
+      "loss": 3.967,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189275748953559e-05,
+      "loss": 3.9843,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1884371542025067e-05,
+      "loss": 3.9835,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1876001973318276e-05,
+      "loss": 3.973,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1867632404611485e-05,
+      "loss": 3.9703,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1859246457100965e-05,
+      "loss": 3.9719,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1850860509590445e-05,
+      "loss": 3.9712,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1842474562079925e-05,
+      "loss": 3.9657,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1834088614569405e-05,
+      "loss": 3.97,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1825702667058885e-05,
+      "loss": 3.9786,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1817316719548365e-05,
+      "loss": 3.9838,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180893077203785e-05,
+      "loss": 3.9761,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1800544824527324e-05,
+      "loss": 3.9588,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1792158877016804e-05,
+      "loss": 3.9734,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178378930831002e-05,
+      "loss": 3.9739,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.17754033607995e-05,
+      "loss": 3.9702,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1767017413288973e-05,
+      "loss": 3.968,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.175863146577845e-05,
+      "loss": 3.9735,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175024551826793e-05,
+      "loss": 3.975,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174185957075741e-05,
+      "loss": 3.9674,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173347362324689e-05,
+      "loss": 3.9706,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172508767573637e-05,
+      "loss": 3.9744,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171670172822585e-05,
+      "loss": 3.9736,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170833215951907e-05,
+      "loss": 3.9811,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169994621200855e-05,
+      "loss": 3.9591,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169156026449803e-05,
+      "loss": 3.9708,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168317431698751e-05,
+      "loss": 3.97,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167480474828072e-05,
+      "loss": 3.9669,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16664188007702e-05,
+      "loss": 3.9892,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165803285325968e-05,
+      "loss": 3.9669,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164964690574916e-05,
+      "loss": 3.9629,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164127733704237e-05,
+      "loss": 3.9665,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163289138953185e-05,
+      "loss": 3.9671,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1624521820825056e-05,
+      "loss": 3.9531,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161615225211827e-05,
+      "loss": 3.964,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160776630460775e-05,
+      "loss": 3.9658,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159938035709723e-05,
+      "loss": 3.9629,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159099440958671e-05,
+      "loss": 3.9617,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158260846207619e-05,
+      "loss": 3.967,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157422251456567e-05,
+      "loss": 3.9581,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156583656705515e-05,
+      "loss": 3.9786,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155745061954463e-05,
+      "loss": 3.9673,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154906467203411e-05,
+      "loss": 3.9737,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154069510332732e-05,
+      "loss": 3.9758,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15323091558168e-05,
+      "loss": 3.9888,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152393958711001e-05,
+      "loss": 3.9699,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1515553639599496e-05,
+      "loss": 3.9619,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1507167692088976e-05,
+      "loss": 3.9736,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498781744578456e-05,
+      "loss": 3.9674,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490395797067936e-05,
+      "loss": 3.9704,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1482009849557416e-05,
+      "loss": 3.9738,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1473623902046896e-05,
+      "loss": 3.9668,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1465237954536376e-05,
+      "loss": 3.9776,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1456852007025856e-05,
+      "loss": 3.9683,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1448466059515336e-05,
+      "loss": 3.9639,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1440096490808545e-05,
+      "loss": 3.9678,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1431710543298025e-05,
+      "loss": 3.9727,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1423324595787505e-05,
+      "loss": 3.9686,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141493864827698e-05,
+      "loss": 3.9748,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1406552700766465e-05,
+      "loss": 3.9704,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139818313205968e-05,
+      "loss": 3.9626,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138979718454916e-05,
+      "loss": 3.9712,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1381411237038634e-05,
+      "loss": 3.972,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1373025289528114e-05,
+      "loss": 3.9756,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1364639342017594e-05,
+      "loss": 3.9674,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135626977331081e-05,
+      "loss": 3.971,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134788382580028e-05,
+      "loss": 3.978,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133949787828976e-05,
+      "loss": 3.9806,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133111193077924e-05,
+      "loss": 3.966,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132272598326872e-05,
+      "loss": 3.9684,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131435641456193e-05,
+      "loss": 3.9631,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130597046705142e-05,
+      "loss": 3.9697,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12975845195409e-05,
+      "loss": 3.9584,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128919857203038e-05,
+      "loss": 3.9763,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128082900332359e-05,
+      "loss": 3.968,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1272459434616803e-05,
+      "loss": 3.973,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1264073487106283e-05,
+      "loss": 3.9668,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.023587226867676,
+      "eval_runtime": 302.7951,
+      "eval_samples_per_second": 1260.228,
+      "eval_steps_per_second": 39.383,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125568753959576e-05,
+      "loss": 3.968,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1247301592085237e-05,
+      "loss": 3.9619,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1238915644574717e-05,
+      "loss": 3.9693,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1230546075867926e-05,
+      "loss": 3.9715,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1222160128357406e-05,
+      "loss": 3.9671,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1213774180846886e-05,
+      "loss": 3.9699,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120538823333637e-05,
+      "loss": 3.9637,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119700228582585e-05,
+      "loss": 3.965,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118861633831533e-05,
+      "loss": 3.9651,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118023039080481e-05,
+      "loss": 3.9671,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117184444329429e-05,
+      "loss": 3.9653,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116345849578377e-05,
+      "loss": 3.9782,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115508892707698e-05,
+      "loss": 3.9802,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114670297956646e-05,
+      "loss": 3.9591,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113831703205594e-05,
+      "loss": 3.969,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112993108454542e-05,
+      "loss": 3.9562,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11215451370349e-05,
+      "loss": 3.9666,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111317556832811e-05,
+      "loss": 3.9593,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110478962081759e-05,
+      "loss": 3.9599,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109640367330707e-05,
+      "loss": 3.9639,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108801772579655e-05,
+      "loss": 3.9781,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107963177828604e-05,
+      "loss": 3.9699,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107124583077552e-05,
+      "loss": 3.9715,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1062859883264997e-05,
+      "loss": 3.9746,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1054490314558206e-05,
+      "loss": 3.9689,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1046120745851415e-05,
+      "loss": 3.9645,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1037734798340895e-05,
+      "loss": 3.9584,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1029348850830375e-05,
+      "loss": 3.9636,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1020962903319855e-05,
+      "loss": 3.961,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1012576955809335e-05,
+      "loss": 3.9633,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1004191008298815e-05,
+      "loss": 3.9677,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0995821439592024e-05,
+      "loss": 3.9596,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098745187088524e-05,
+      "loss": 3.9696,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097906592337472e-05,
+      "loss": 3.9639,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09706799758642e-05,
+      "loss": 3.964,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096231040715741e-05,
+      "loss": 3.9759,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095392445964689e-05,
+      "loss": 3.9681,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094553851213637e-05,
+      "loss": 3.9449,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093715256462585e-05,
+      "loss": 3.9708,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092876661711533e-05,
+      "loss": 3.9547,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092038066960481e-05,
+      "loss": 3.9472,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091199472209429e-05,
+      "loss": 3.968,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090360877458376e-05,
+      "loss": 3.974,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089522282707324e-05,
+      "loss": 3.962,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088683687956273e-05,
+      "loss": 3.9653,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087845093205221e-05,
+      "loss": 3.9723,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087006498454169e-05,
+      "loss": 3.9672,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086167903703117e-05,
+      "loss": 3.9636,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085329308952065e-05,
+      "loss": 3.9466,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084490714201013e-05,
+      "loss": 3.9652,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083652119449961e-05,
+      "loss": 3.9688,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0828168004596546e-05,
+      "loss": 3.9603,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0819782057086026e-05,
+      "loss": 3.9559,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0811396109575506e-05,
+      "loss": 3.953,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0803010162064986e-05,
+      "loss": 3.9578,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0794624214554466e-05,
+      "loss": 3.9514,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0786238267043946e-05,
+      "loss": 3.9723,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0777852319533426e-05,
+      "loss": 3.9496,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076946637202291e-05,
+      "loss": 3.9701,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076108042451239e-05,
+      "loss": 3.9649,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07527108558056e-05,
+      "loss": 3.9473,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074432490829508e-05,
+      "loss": 3.9593,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073593896078456e-05,
+      "loss": 3.957,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072755301327404e-05,
+      "loss": 3.9534,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071918344456725e-05,
+      "loss": 3.9497,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071079749705673e-05,
+      "loss": 3.9759,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070241154954621e-05,
+      "loss": 3.9581,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069404198083942e-05,
+      "loss": 3.9532,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06856560333289e-05,
+      "loss": 3.9417,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067727008581838e-05,
+      "loss": 3.9532,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0668884138307866e-05,
+      "loss": 3.9693,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0660498190797346e-05,
+      "loss": 3.972,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0652112243286826e-05,
+      "loss": 3.9579,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0643742674580035e-05,
+      "loss": 3.9706,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0635356727069515e-05,
+      "loss": 3.9709,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0626970779558995e-05,
+      "loss": 3.9611,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0618584832048475e-05,
+      "loss": 3.9586,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061019888453795e-05,
+      "loss": 3.9633,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060181293702743e-05,
+      "loss": 3.9577,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059342698951691e-05,
+      "loss": 3.9565,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058504104200639e-05,
+      "loss": 3.9557,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0576671473299604e-05,
+      "loss": 3.9662,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056830190459281e-05,
+      "loss": 3.9689,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05599159570823e-05,
+      "loss": 3.9668,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055153000957178e-05,
+      "loss": 3.9483,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054314406206125e-05,
+      "loss": 3.9619,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053475811455073e-05,
+      "loss": 3.9645,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052637216704021e-05,
+      "loss": 3.9613,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051798621952969e-05,
+      "loss": 3.9576,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509600272019173e-05,
+      "loss": 3.9601,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501230703312382e-05,
+      "loss": 3.967,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492844755801865e-05,
+      "loss": 3.9545,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484458808291345e-05,
+      "loss": 3.9578,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0476072860780825e-05,
+      "loss": 3.9625,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0467686913270305e-05,
+      "loss": 3.9608,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0459317344563514e-05,
+      "loss": 3.969,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0450931397052994e-05,
+      "loss": 3.9554,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0442545449542474e-05,
+      "loss": 3.9544,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0434159502031957e-05,
+      "loss": 3.9598,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0425789933325167e-05,
+      "loss": 3.9599,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0417403985814647e-05,
+      "loss": 3.9722,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0409018038304126e-05,
+      "loss": 3.9594,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0400632090793606e-05,
+      "loss": 3.9509,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.039226252208682e-05,
+      "loss": 3.9525,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383892953380028e-05,
+      "loss": 3.9636,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0375507005869508e-05,
+      "loss": 3.9359,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0367121058358988e-05,
+      "loss": 3.9555,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0358735110848468e-05,
+      "loss": 3.9586,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0350349163337948e-05,
+      "loss": 3.9504,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0341963215827428e-05,
+      "loss": 3.9508,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.033359364712064e-05,
+      "loss": 3.9593,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.032520769961012e-05,
+      "loss": 3.9446,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03168217520996e-05,
+      "loss": 3.967,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.030843580458908e-05,
+      "loss": 3.955,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.030004985707856e-05,
+      "loss": 3.9608,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.029166390956804e-05,
+      "loss": 3.9654,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.028327796205752e-05,
+      "loss": 3.9773,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0274908393350732e-05,
+      "loss": 3.9618,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266522445840212e-05,
+      "loss": 3.9488,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0258136498329692e-05,
+      "loss": 3.9627,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249750550819172e-05,
+      "loss": 3.9548,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241364603308652e-05,
+      "loss": 3.9589,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.023299503460186e-05,
+      "loss": 3.9611,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224625465895074e-05,
+      "loss": 3.958,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216239518384554e-05,
+      "loss": 3.9695,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0207853570874034e-05,
+      "loss": 3.9566,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0199467623363514e-05,
+      "loss": 3.9521,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0191081675852994e-05,
+      "loss": 3.9555,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182695728342474e-05,
+      "loss": 3.9588,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174309780831954e-05,
+      "loss": 3.9539,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0165923833321437e-05,
+      "loss": 3.9635,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015753788581091e-05,
+      "loss": 3.9647,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0149168317104126e-05,
+      "loss": 3.9565,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140782369593606e-05,
+      "loss": 3.9506,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132396422083083e-05,
+      "loss": 3.9611,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0124010474572563e-05,
+      "loss": 3.9634,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115624527062042e-05,
+      "loss": 3.9602,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107238579551522e-05,
+      "loss": 3.9607,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098852632041002e-05,
+      "loss": 3.9678,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009048306333421e-05,
+      "loss": 3.9664,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.008209711582369e-05,
+      "loss": 3.9568,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073711168313175e-05,
+      "loss": 3.9533,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065325220802655e-05,
+      "loss": 3.9544,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056939273292135e-05,
+      "loss": 3.9599,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048553325781615e-05,
+      "loss": 3.9445,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0040167378271095e-05,
+      "loss": 3.9653,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0031781430760575e-05,
+      "loss": 3.9629,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0023411862053784e-05,
+      "loss": 3.9605,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0015025914543267e-05,
+      "loss": 3.955,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0006639967032747e-05,
+      "loss": 3.9616,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.01820707321167,
+      "eval_runtime": 294.1708,
+      "eval_samples_per_second": 1297.175,
+      "eval_steps_per_second": 40.538,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998254019522227e-05,
+      "loss": 3.9471,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989868072011707e-05,
+      "loss": 3.9544,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981482124501187e-05,
+      "loss": 3.9614,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973096176990667e-05,
+      "loss": 3.9586,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964710229480147e-05,
+      "loss": 3.9615,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956324281969623e-05,
+      "loss": 3.9517,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947938334459103e-05,
+      "loss": 3.955,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939552386948583e-05,
+      "loss": 3.949,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931166439438063e-05,
+      "loss": 3.9633,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922780491927543e-05,
+      "loss": 3.9488,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914394544417023e-05,
+      "loss": 3.9706,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906024975710236e-05,
+      "loss": 3.971,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897639028199716e-05,
+      "loss": 3.9462,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889253080689195e-05,
+      "loss": 3.9546,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880867133178675e-05,
+      "loss": 3.9454,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872481185668155e-05,
+      "loss": 3.9547,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864095238157635e-05,
+      "loss": 3.9517,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855709290647115e-05,
+      "loss": 3.9491,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847323343136595e-05,
+      "loss": 3.9526,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9838937395626075e-05,
+      "loss": 3.9646,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830567826919288e-05,
+      "loss": 3.9584,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822181879408768e-05,
+      "loss": 3.9609,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813795931898248e-05,
+      "loss": 3.9641,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805409984387728e-05,
+      "loss": 3.965,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9797040415680937e-05,
+      "loss": 3.9539,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788654468170417e-05,
+      "loss": 3.9479,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.97802685206599e-05,
+      "loss": 3.953,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.977188257314938e-05,
+      "loss": 3.9494,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976349662563886e-05,
+      "loss": 3.9515,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755110678128333e-05,
+      "loss": 3.9524,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746724730617813e-05,
+      "loss": 3.9552,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9738338783107293e-05,
+      "loss": 3.9592,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972996921440051e-05,
+      "loss": 3.9497,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721583266889985e-05,
+      "loss": 3.9557,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97132136981832e-05,
+      "loss": 3.9638,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970482775067268e-05,
+      "loss": 3.9576,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696441803162154e-05,
+      "loss": 3.9386,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9688055855651638e-05,
+      "loss": 3.9584,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679686286944854e-05,
+      "loss": 3.9403,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9671300339434334e-05,
+      "loss": 3.944,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662914391923807e-05,
+      "loss": 3.954,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654528444413287e-05,
+      "loss": 3.9604,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646158875706503e-05,
+      "loss": 3.9521,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637772928195983e-05,
+      "loss": 3.9555,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962938698068546e-05,
+      "loss": 3.9602,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962100103317494e-05,
+      "loss": 3.9556,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961261508566442e-05,
+      "loss": 3.9548,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96042291381539e-05,
+      "loss": 3.9379,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959584319064338e-05,
+      "loss": 3.9499,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958745724313286e-05,
+      "loss": 3.9595,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957908767442607e-05,
+      "loss": 3.9484,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957070172691555e-05,
+      "loss": 3.9444,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.956233215820876e-05,
+      "loss": 3.9426,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.955394621069824e-05,
+      "loss": 3.9497,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.954556026318772e-05,
+      "loss": 3.9368,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537190694480933e-05,
+      "loss": 3.9595,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528804746970413e-05,
+      "loss": 3.9422,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520418799459893e-05,
+      "loss": 3.9583,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512032851949373e-05,
+      "loss": 3.9523,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503646904438853e-05,
+      "loss": 3.9346,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495260956928333e-05,
+      "loss": 3.9531,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486875009417813e-05,
+      "loss": 3.9464,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478489061907292e-05,
+      "loss": 3.9413,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9470119493200505e-05,
+      "loss": 3.9411,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461733545689985e-05,
+      "loss": 3.9631,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453347598179465e-05,
+      "loss": 3.9505,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9444961650668945e-05,
+      "loss": 3.9405,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436575703158425e-05,
+      "loss": 3.9322,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9428206134451637e-05,
+      "loss": 3.9401,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419820186941117e-05,
+      "loss": 3.9571,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411434239430597e-05,
+      "loss": 3.9628,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403048291920077e-05,
+      "loss": 3.9452,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394678723213286e-05,
+      "loss": 3.9607,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386292775702766e-05,
+      "loss": 3.9614,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377906828192246e-05,
+      "loss": 3.9509,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936952088068173e-05,
+      "loss": 3.943,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936113493317121e-05,
+      "loss": 3.9522,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935274898566069e-05,
+      "loss": 3.9483,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93443794169539e-05,
+      "loss": 3.9464,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.933599346944338e-05,
+      "loss": 3.9472,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932760752193286e-05,
+      "loss": 3.9495,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931922157442234e-05,
+      "loss": 3.9589,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310835626911815e-05,
+      "loss": 3.9545,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302449679401295e-05,
+      "loss": 3.9419,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294063731890775e-05,
+      "loss": 3.9495,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285677784380255e-05,
+      "loss": 3.9532,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9277291836869735e-05,
+      "loss": 3.9482,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9268905889359215e-05,
+      "loss": 3.9476,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260519941848698e-05,
+      "loss": 3.9469,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252133994338178e-05,
+      "loss": 3.9576,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243780804435116e-05,
+      "loss": 3.9458,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235394856924596e-05,
+      "loss": 3.9482,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9227008909414076e-05,
+      "loss": 3.9485,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921862296190356e-05,
+      "loss": 3.9535,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921023701439304e-05,
+      "loss": 3.9528,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920185106688252e-05,
+      "loss": 3.9506,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.919348149817573e-05,
+      "loss": 3.9379,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.918509555066521e-05,
+      "loss": 3.9494,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917670960315469e-05,
+      "loss": 3.9504,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916832365564417e-05,
+      "loss": 3.9616,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915993770813365e-05,
+      "loss": 3.9503,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915155176062313e-05,
+      "loss": 3.9423,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.914316581311261e-05,
+      "loss": 3.9409,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.913477986560209e-05,
+      "loss": 3.9552,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91264102968953e-05,
+      "loss": 3.9268,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.911802434938478e-05,
+      "loss": 3.9396,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109654780677993e-05,
+      "loss": 3.9454,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101268833167473e-05,
+      "loss": 3.9449,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092882885656953e-05,
+      "loss": 3.9361,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084496938146433e-05,
+      "loss": 3.9494,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076110990635913e-05,
+      "loss": 3.9324,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067725043125393e-05,
+      "loss": 3.9597,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059339095614873e-05,
+      "loss": 3.9424,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050953148104353e-05,
+      "loss": 3.9511,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904256720059383e-05,
+      "loss": 3.9555,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034197631887045e-05,
+      "loss": 3.969,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025828063180254e-05,
+      "loss": 3.9452,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017442115669734e-05,
+      "loss": 3.9444,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9009056168159214e-05,
+      "loss": 3.9505,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000670220648694e-05,
+      "loss": 3.9419,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992284273138178e-05,
+      "loss": 3.952,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898389832562765e-05,
+      "loss": 3.95,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897551237811713e-05,
+      "loss": 3.9432,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967142809410347e-05,
+      "loss": 3.9617,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958756861899827e-05,
+      "loss": 3.9441,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950370914389303e-05,
+      "loss": 3.9422,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8941984966878783e-05,
+      "loss": 3.9443,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933615398172e-05,
+      "loss": 3.9501,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.892522945066148e-05,
+      "loss": 3.941,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916843503150952e-05,
+      "loss": 3.9599,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908457555640432e-05,
+      "loss": 3.9559,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900071608129915e-05,
+      "loss": 3.9447,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891685660619395e-05,
+      "loss": 3.9405,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883299713108875e-05,
+      "loss": 3.9497,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8874913765598355e-05,
+      "loss": 3.9503,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8866544196891564e-05,
+      "loss": 3.9531,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858158249381044e-05,
+      "loss": 3.9493,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849788680674257e-05,
+      "loss": 3.9585,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841402733163737e-05,
+      "loss": 3.9531,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8833016785653217e-05,
+      "loss": 3.9474,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8824630838142697e-05,
+      "loss": 3.9463,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8816261269435906e-05,
+      "loss": 3.9442,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807875321925386e-05,
+      "loss": 3.9516,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879948937441487e-05,
+      "loss": 3.9314,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879110342690435e-05,
+      "loss": 3.955,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.878271747939383e-05,
+      "loss": 3.9535,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774347910687038e-05,
+      "loss": 3.9483,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765961963176518e-05,
+      "loss": 3.9445,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757576015665998e-05,
+      "loss": 3.9493,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.012906551361084,
+      "eval_runtime": 294.1322,
+      "eval_samples_per_second": 1297.345,
+      "eval_steps_per_second": 40.543,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8749190068155478e-05,
+      "loss": 3.9365,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874082049944869e-05,
+      "loss": 3.944,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.873243455193817e-05,
+      "loss": 3.9552,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.872404860442765e-05,
+      "loss": 3.9441,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.871566265691713e-05,
+      "loss": 3.951,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.870727670940661e-05,
+      "loss": 3.9451,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869889076189609e-05,
+      "loss": 3.9434,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869050481438557e-05,
+      "loss": 3.9414,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8682135245678783e-05,
+      "loss": 3.9512,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8673749298168263e-05,
+      "loss": 3.9405,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8665363350657743e-05,
+      "loss": 3.9604,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8656977403147222e-05,
+      "loss": 3.9568,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.864860783444043e-05,
+      "loss": 3.9434,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640221886929915e-05,
+      "loss": 3.9399,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631835939419395e-05,
+      "loss": 3.9406,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623449991908875e-05,
+      "loss": 3.9395,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8615064044398355e-05,
+      "loss": 3.9397,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8606678096887835e-05,
+      "loss": 3.9408,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8598292149377308e-05,
+      "loss": 3.9384,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858990620186679e-05,
+      "loss": 3.9585,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858152025435627e-05,
+      "loss": 3.9476,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8573150685649487e-05,
+      "loss": 3.9528,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564781116942696e-05,
+      "loss": 3.951,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556395169432176e-05,
+      "loss": 3.9573,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8548009221921656e-05,
+      "loss": 3.9361,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539623274411136e-05,
+      "loss": 3.9417,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531237326900613e-05,
+      "loss": 3.9378,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522851379390093e-05,
+      "loss": 3.9425,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514465431879573e-05,
+      "loss": 3.9422,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8506079484369052e-05,
+      "loss": 3.9418,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497693536858532e-05,
+      "loss": 3.9465,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489307589348012e-05,
+      "loss": 3.947,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8480921641837492e-05,
+      "loss": 3.9373,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472552073130705e-05,
+      "loss": 3.95,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464182504423914e-05,
+      "loss": 3.9539,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455796556913394e-05,
+      "loss": 3.9432,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447410609402874e-05,
+      "loss": 3.9325,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8439024661892354e-05,
+      "loss": 3.9458,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8430638714381834e-05,
+      "loss": 3.9293,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422252766871317e-05,
+      "loss": 3.9382,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8413866819360797e-05,
+      "loss": 3.9417,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405480871850277e-05,
+      "loss": 3.9519,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397094924339757e-05,
+      "loss": 3.9411,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388708976829237e-05,
+      "loss": 3.9455,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380323029318717e-05,
+      "loss": 3.9522,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8371937081808197e-05,
+      "loss": 3.945,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.836356751310141e-05,
+      "loss": 3.946,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835518156559089e-05,
+      "loss": 3.9294,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.834679561808037e-05,
+      "loss": 3.9418,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.833840967056985e-05,
+      "loss": 3.9481,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8330023723059322e-05,
+      "loss": 3.9383,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321637775548802e-05,
+      "loss": 3.9378,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313251828038286e-05,
+      "loss": 3.9275,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304865880527766e-05,
+      "loss": 3.9395,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296496311820975e-05,
+      "loss": 3.9289,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.828812674311419e-05,
+      "loss": 3.9462,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.827974079560367e-05,
+      "loss": 3.9353,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271354848093147e-05,
+      "loss": 3.9458,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8262968900582627e-05,
+      "loss": 3.9471,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254582953072107e-05,
+      "loss": 3.9242,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8246197005561587e-05,
+      "loss": 3.938,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8237827436854796e-05,
+      "loss": 3.9414,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229441489344276e-05,
+      "loss": 3.9293,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8221055541833756e-05,
+      "loss": 3.932,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.821266959432324e-05,
+      "loss": 3.9518,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.820428364681272e-05,
+      "loss": 3.942,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81958976993022e-05,
+      "loss": 3.9304,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818751175179168e-05,
+      "loss": 3.924,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817912580428116e-05,
+      "loss": 3.9341,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8170756235574368e-05,
+      "loss": 3.9458,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.816238666686758e-05,
+      "loss": 3.9513,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.815400071935706e-05,
+      "loss": 3.9383,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.814561477184654e-05,
+      "loss": 3.9509,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.813722882433602e-05,
+      "loss": 3.9479,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81288428768255e-05,
+      "loss": 3.9442,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.812047330811871e-05,
+      "loss": 3.9315,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8112087360608193e-05,
+      "loss": 3.9473,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103701413097673e-05,
+      "loss": 3.9393,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095315465587153e-05,
+      "loss": 3.9358,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086929518076633e-05,
+      "loss": 3.9392,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078543570566113e-05,
+      "loss": 3.936,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070157623055593e-05,
+      "loss": 3.9554,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061788054348802e-05,
+      "loss": 3.9399,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053402106838282e-05,
+      "loss": 3.9349,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8045016159327765e-05,
+      "loss": 3.9403,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8036630211817245e-05,
+      "loss": 3.9424,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028244264306725e-05,
+      "loss": 3.9386,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8019858316796205e-05,
+      "loss": 3.9402,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8011488748089414e-05,
+      "loss": 3.9326,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8003102800578894e-05,
+      "loss": 3.9501,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994716853068374e-05,
+      "loss": 3.9341,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7986330905557857e-05,
+      "loss": 3.9378,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7977961336851066e-05,
+      "loss": 3.9341,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7969575389340546e-05,
+      "loss": 3.9477,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961189441830026e-05,
+      "loss": 3.9455,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952803494319506e-05,
+      "loss": 3.9422,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7944417546808983e-05,
+      "loss": 3.9298,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7936031599298463e-05,
+      "loss": 3.9398,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927645651787943e-05,
+      "loss": 3.941,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919259704277423e-05,
+      "loss": 3.9514,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910873756766903e-05,
+      "loss": 3.9422,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902520566863848e-05,
+      "loss": 3.9348,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894134619353328e-05,
+      "loss": 3.9295,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885748671842804e-05,
+      "loss": 3.9442,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877362724332284e-05,
+      "loss": 3.9159,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7868976776821764e-05,
+      "loss": 3.9326,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860590829311244e-05,
+      "loss": 3.9382,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7852204881800724e-05,
+      "loss": 3.9363,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843818934290204e-05,
+      "loss": 3.9253,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835449365583417e-05,
+      "loss": 3.9368,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7827079796876632e-05,
+      "loss": 3.9284,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7818693849366106e-05,
+      "loss": 3.9465,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810307901855586e-05,
+      "loss": 3.9343,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801921954345066e-05,
+      "loss": 3.9358,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.779353600683455e-05,
+      "loss": 3.9518,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785166438127758e-05,
+      "loss": 3.9569,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776796869420974e-05,
+      "loss": 3.9362,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768410921910454e-05,
+      "loss": 3.9321,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7760024974399927e-05,
+      "loss": 3.943,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.775163902688941e-05,
+      "loss": 3.9341,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.774325307937889e-05,
+      "loss": 3.938,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.773486713186837e-05,
+      "loss": 3.9443,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.772648118435785e-05,
+      "loss": 3.9328,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.771809523684733e-05,
+      "loss": 3.9514,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770970928933681e-05,
+      "loss": 3.9333,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770132334182629e-05,
+      "loss": 3.9358,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.769293739431577e-05,
+      "loss": 3.9305,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.768455144680525e-05,
+      "loss": 3.9393,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676181878098462e-05,
+      "loss": 3.9346,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667795930587942e-05,
+      "loss": 3.9484,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659409983077422e-05,
+      "loss": 3.9469,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765104041437063e-05,
+      "loss": 3.9386,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.764265446686011e-05,
+      "loss": 3.9262,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634268519349595e-05,
+      "loss": 3.9452,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625882571839075e-05,
+      "loss": 3.9365,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617513003132284e-05,
+      "loss": 3.945,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609143434425493e-05,
+      "loss": 3.9389,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7600757486914973e-05,
+      "loss": 3.9447,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7592371539404456e-05,
+      "loss": 3.9421,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7583985591893936e-05,
+      "loss": 3.9463,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7575599644383416e-05,
+      "loss": 3.9342,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7567213696872896e-05,
+      "loss": 3.9332,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558827749362376e-05,
+      "loss": 3.9418,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7550441801851856e-05,
+      "loss": 3.9267,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7542055854341336e-05,
+      "loss": 3.9408,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7533669906830816e-05,
+      "loss": 3.9494,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525283959320292e-05,
+      "loss": 3.9377,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7516898011809772e-05,
+      "loss": 3.9332,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508528443102988e-05,
+      "loss": 3.9384,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.007848739624023,
+      "eval_runtime": 297.5592,
+      "eval_samples_per_second": 1282.404,
+      "eval_steps_per_second": 40.076,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.750014249559246e-05,
+      "loss": 3.9278,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.749175654808194e-05,
+      "loss": 3.933,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483370600571425e-05,
+      "loss": 3.9425,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7474984653060905e-05,
+      "loss": 3.9366,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466598705550385e-05,
+      "loss": 3.9421,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458212758039865e-05,
+      "loss": 3.9297,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449826810529345e-05,
+      "loss": 3.9401,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7441440863018825e-05,
+      "loss": 3.9257,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7433054915508304e-05,
+      "loss": 3.9411,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7424668967997784e-05,
+      "loss": 3.9336,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7416283020487264e-05,
+      "loss": 3.9459,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7407897072976744e-05,
+      "loss": 3.9495,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7399527504269957e-05,
+      "loss": 3.9333,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391141556759437e-05,
+      "loss": 3.9325,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382755609248917e-05,
+      "loss": 3.9345,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374369661738397e-05,
+      "loss": 3.9299,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7365983714227877e-05,
+      "loss": 3.9289,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357597766717353e-05,
+      "loss": 3.9315,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349211819206833e-05,
+      "loss": 3.9276,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7340825871696313e-05,
+      "loss": 3.953,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332439924185793e-05,
+      "loss": 3.933,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324053976675273e-05,
+      "loss": 3.947,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7315668029164753e-05,
+      "loss": 3.9431,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7307298460457965e-05,
+      "loss": 3.9478,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7298912512947445e-05,
+      "loss": 3.922,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290526565436925e-05,
+      "loss": 3.9362,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282140617926405e-05,
+      "loss": 3.926,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273754670415885e-05,
+      "loss": 3.9365,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265385101709094e-05,
+      "loss": 3.9307,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7256999154198574e-05,
+      "loss": 3.9328,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7248613206688058e-05,
+      "loss": 3.9415,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7240227259177538e-05,
+      "loss": 3.9366,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7231857690470747e-05,
+      "loss": 3.9264,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223471742960227e-05,
+      "loss": 3.9422,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7215102174253436e-05,
+      "loss": 3.938,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720671622674292e-05,
+      "loss": 3.9327,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.71983302792324e-05,
+      "loss": 3.9244,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718994433172188e-05,
+      "loss": 3.9379,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718155838421136e-05,
+      "loss": 3.9187,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173188815504568e-05,
+      "loss": 3.932,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164802867994048e-05,
+      "loss": 3.9314,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156416920483528e-05,
+      "loss": 3.9399,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.714803097297301e-05,
+      "loss": 3.9337,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.713964502546249e-05,
+      "loss": 3.934,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.713125907795197e-05,
+      "loss": 3.9417,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.712287313044145e-05,
+      "loss": 3.9381,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.711448718293093e-05,
+      "loss": 3.9403,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.710611761422414e-05,
+      "loss": 3.9197,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709773166671362e-05,
+      "loss": 3.9281,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089345719203104e-05,
+      "loss": 3.9399,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7080976150496313e-05,
+      "loss": 3.9298,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072590202985793e-05,
+      "loss": 3.9296,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064220634279002e-05,
+      "loss": 3.9217,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7055834686768482e-05,
+      "loss": 3.9308,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7047448739257965e-05,
+      "loss": 3.9173,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7039062791747445e-05,
+      "loss": 3.9352,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030676844236925e-05,
+      "loss": 3.9268,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022290896726405e-05,
+      "loss": 3.9394,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7013904949215885e-05,
+      "loss": 3.936,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7005519001705365e-05,
+      "loss": 3.9185,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.699713305419484e-05,
+      "loss": 3.9278,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698874710668432e-05,
+      "loss": 3.9313,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69803611591738e-05,
+      "loss": 3.9192,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697197521166328e-05,
+      "loss": 3.931,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.696360564295649e-05,
+      "loss": 3.9368,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.695521969544597e-05,
+      "loss": 3.9359,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.694683374793545e-05,
+      "loss": 3.9225,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693844780042493e-05,
+      "loss": 3.9148,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930078231718143e-05,
+      "loss": 3.9204,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921692284207623e-05,
+      "loss": 3.935,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913306336697103e-05,
+      "loss": 3.9431,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6904936767990312e-05,
+      "loss": 3.9291,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689655082047979e-05,
+      "loss": 3.9426,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888164872969275e-05,
+      "loss": 3.9406,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879778925458755e-05,
+      "loss": 3.9395,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871409356751964e-05,
+      "loss": 3.9175,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6863023409241444e-05,
+      "loss": 3.9375,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854637461730924e-05,
+      "loss": 3.9319,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6846251514220404e-05,
+      "loss": 3.9255,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6837865566709884e-05,
+      "loss": 3.9319,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6829479619199367e-05,
+      "loss": 3.9283,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821093671688847e-05,
+      "loss": 3.9415,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812707724178327e-05,
+      "loss": 3.9338,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804338155471536e-05,
+      "loss": 3.9232,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6795952207961016e-05,
+      "loss": 3.9315,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787566260450496e-05,
+      "loss": 3.9317,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779180312939976e-05,
+      "loss": 3.9312,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.677081074423319e-05,
+      "loss": 3.931,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.676242479672267e-05,
+      "loss": 3.926,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6754055228015878e-05,
+      "loss": 3.9374,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745669280505358e-05,
+      "loss": 3.9267,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737283332994838e-05,
+      "loss": 3.9308,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.672889738548432e-05,
+      "loss": 3.924,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.67205114379738e-05,
+      "loss": 3.9404,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.671212549046328e-05,
+      "loss": 3.9342,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.670373954295276e-05,
+      "loss": 3.9312,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.669535359544224e-05,
+      "loss": 3.9241,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.668696764793172e-05,
+      "loss": 3.9325,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.66785817004212e-05,
+      "loss": 3.924,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670195752910677e-05,
+      "loss": 3.9434,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661809805400157e-05,
+      "loss": 3.93,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653440236693373e-05,
+      "loss": 3.9272,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645054289182846e-05,
+      "loss": 3.9205,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636668341672326e-05,
+      "loss": 3.9322,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628282394161806e-05,
+      "loss": 3.9102,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6619912825455022e-05,
+      "loss": 3.9247,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.66115268779445e-05,
+      "loss": 3.925,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.660314093043398e-05,
+      "loss": 3.9286,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6594771361727194e-05,
+      "loss": 3.9128,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6586385414216674e-05,
+      "loss": 3.9319,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657799946670615e-05,
+      "loss": 3.9174,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656961351919563e-05,
+      "loss": 3.9402,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656122757168511e-05,
+      "loss": 3.9257,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.655284162417459e-05,
+      "loss": 3.9242,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.654445567666407e-05,
+      "loss": 3.9431,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.653606972915355e-05,
+      "loss": 3.9501,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652770016044676e-05,
+      "loss": 3.93,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519314212936243e-05,
+      "loss": 3.9214,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6510928265425723e-05,
+      "loss": 3.9315,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502542317915203e-05,
+      "loss": 3.9255,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494156370404683e-05,
+      "loss": 3.9279,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648580318050162e-05,
+      "loss": 3.9409,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477417232991105e-05,
+      "loss": 3.9201,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6469031285480585e-05,
+      "loss": 3.9424,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460645337970064e-05,
+      "loss": 3.9277,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452259390459544e-05,
+      "loss": 3.929,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6443873442949024e-05,
+      "loss": 3.921,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6435487495438504e-05,
+      "loss": 3.9321,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6427101547927984e-05,
+      "loss": 3.922,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6418731979221197e-05,
+      "loss": 3.9403,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6410346031710677e-05,
+      "loss": 3.933,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6401960084200157e-05,
+      "loss": 3.932,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6393574136689637e-05,
+      "loss": 3.9178,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385188189179117e-05,
+      "loss": 3.936,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376818620472326e-05,
+      "loss": 3.9332,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368432672961806e-05,
+      "loss": 3.9314,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360046725451286e-05,
+      "loss": 3.9281,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.635166077794077e-05,
+      "loss": 3.9386,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.634327483043025e-05,
+      "loss": 3.9344,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.633488888291973e-05,
+      "loss": 3.9379,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.632650293540921e-05,
+      "loss": 3.9226,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318116987898682e-05,
+      "loss": 3.9229,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309747419191898e-05,
+      "loss": 3.931,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301361471681378e-05,
+      "loss": 3.9168,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.629297552417086e-05,
+      "loss": 3.9307,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284589576660334e-05,
+      "loss": 3.9404,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276203629149814e-05,
+      "loss": 3.927,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.626783406044303e-05,
+      "loss": 3.9222,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259448112932507e-05,
+      "loss": 3.9327,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.003542900085449,
+      "eval_runtime": 295.5844,
+      "eval_samples_per_second": 1290.972,
+      "eval_steps_per_second": 40.344,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6251062165421987e-05,
+      "loss": 3.9269,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242676217911467e-05,
+      "loss": 3.9274,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234306649204683e-05,
+      "loss": 3.9341,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225920701694156e-05,
+      "loss": 3.9235,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217534754183636e-05,
+      "loss": 3.9355,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209148806673116e-05,
+      "loss": 3.9209,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.620077923796633e-05,
+      "loss": 3.9318,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192393290455808e-05,
+      "loss": 3.9165,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6184023721749024e-05,
+      "loss": 3.9322,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6175637774238504e-05,
+      "loss": 3.9234,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167251826727977e-05,
+      "loss": 3.9369,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615886587921746e-05,
+      "loss": 3.9384,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6150496310510676e-05,
+      "loss": 3.9229,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6142110363000156e-05,
+      "loss": 3.9271,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.613372441548963e-05,
+      "loss": 3.9233,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.612533846797911e-05,
+      "loss": 3.9244,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.611695252046859e-05,
+      "loss": 3.9165,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.610856657295807e-05,
+      "loss": 3.9228,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6100180625447553e-05,
+      "loss": 3.9207,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6091794677937033e-05,
+      "loss": 3.9493,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083425109230242e-05,
+      "loss": 3.9249,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.607503916171972e-05,
+      "loss": 3.9319,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.60666532142092e-05,
+      "loss": 3.9376,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.605826726669868e-05,
+      "loss": 3.9356,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.604988131918816e-05,
+      "loss": 3.9169,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041495371677645e-05,
+      "loss": 3.9252,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033109424167125e-05,
+      "loss": 3.9182,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6024723476656605e-05,
+      "loss": 3.9267,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6016353907949814e-05,
+      "loss": 3.9199,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6007967960439294e-05,
+      "loss": 3.9212,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5999582012928774e-05,
+      "loss": 3.932,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5991196065418254e-05,
+      "loss": 3.9277,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5982826496711466e-05,
+      "loss": 3.9243,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5974440549200946e-05,
+      "loss": 3.9326,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966070980494155e-05,
+      "loss": 3.9298,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957685032983635e-05,
+      "loss": 3.9254,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949299085473115e-05,
+      "loss": 3.9118,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.59409131379626e-05,
+      "loss": 3.9316,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.593252719045208e-05,
+      "loss": 3.9087,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.592414124294156e-05,
+      "loss": 3.9246,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5915771674234768e-05,
+      "loss": 3.9218,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907385726724247e-05,
+      "loss": 3.9312,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5898999779213727e-05,
+      "loss": 3.9294,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890613831703207e-05,
+      "loss": 3.9226,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.588222788419269e-05,
+      "loss": 3.9321,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.587384193668217e-05,
+      "loss": 3.9315,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865455989171644e-05,
+      "loss": 3.9283,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5857070041661124e-05,
+      "loss": 3.914,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5848684094150604e-05,
+      "loss": 3.9192,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5840298146640084e-05,
+      "loss": 3.9347,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5831912199129564e-05,
+      "loss": 3.9191,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5823542630422776e-05,
+      "loss": 3.9169,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815156682912256e-05,
+      "loss": 3.9167,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5806770735401736e-05,
+      "loss": 3.9252,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798384787891216e-05,
+      "loss": 3.9123,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5789998840380696e-05,
+      "loss": 3.9234,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781612892870176e-05,
+      "loss": 3.9212,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773226945359656e-05,
+      "loss": 3.9248,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.576484099784914e-05,
+      "loss": 3.9262,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.575647142914235e-05,
+      "loss": 3.9132,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5748101860435557e-05,
+      "loss": 3.9141,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739715912925037e-05,
+      "loss": 3.9274,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731329965414517e-05,
+      "loss": 3.9091,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5722944017904e-05,
+      "loss": 3.9208,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.571455807039348e-05,
+      "loss": 3.9269,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.570617212288296e-05,
+      "loss": 3.9308,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569778617537244e-05,
+      "loss": 3.9107,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568940022786192e-05,
+      "loss": 3.9078,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.56810142803514e-05,
+      "loss": 3.9099,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.567264471164461e-05,
+      "loss": 3.9243,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664275142937822e-05,
+      "loss": 3.933,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5655889195427302e-05,
+      "loss": 3.9211,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5647503247916782e-05,
+      "loss": 3.9308,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5639117300406262e-05,
+      "loss": 3.9336,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630731352895742e-05,
+      "loss": 3.9311,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622361784188954e-05,
+      "loss": 3.9112,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5613975836678434e-05,
+      "loss": 3.9285,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5605589889167914e-05,
+      "loss": 3.9235,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5597203941657394e-05,
+      "loss": 3.9177,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5588834372950603e-05,
+      "loss": 3.922,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5580448425440083e-05,
+      "loss": 3.9215,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5572062477929563e-05,
+      "loss": 3.9288,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5563676530419047e-05,
+      "loss": 3.9265,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5555290582908527e-05,
+      "loss": 3.9171,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5546904635398e-05,
+      "loss": 3.9251,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553851868788748e-05,
+      "loss": 3.9204,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5530149119180696e-05,
+      "loss": 3.923,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5521763171670176e-05,
+      "loss": 3.9218,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513377224159652e-05,
+      "loss": 3.9161,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5504991276649132e-05,
+      "loss": 3.9287,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5496605329138612e-05,
+      "loss": 3.9201,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5488219381628092e-05,
+      "loss": 3.9187,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5479833434117572e-05,
+      "loss": 3.9149,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5471447486607052e-05,
+      "loss": 3.9336,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5463077917900264e-05,
+      "loss": 3.9261,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454691970389744e-05,
+      "loss": 3.9209,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5446306022879224e-05,
+      "loss": 3.9134,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5437920075368704e-05,
+      "loss": 3.9232,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5429550506661913e-05,
+      "loss": 3.9144,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5421164559151393e-05,
+      "loss": 3.9389,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5412778611640877e-05,
+      "loss": 3.9236,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5404392664130357e-05,
+      "loss": 3.92,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5396023095423566e-05,
+      "loss": 3.9142,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5387637147913046e-05,
+      "loss": 3.9246,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5379251200402526e-05,
+      "loss": 3.8983,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5370865252892006e-05,
+      "loss": 3.9172,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5362479305381485e-05,
+      "loss": 3.9157,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.535409335787097e-05,
+      "loss": 3.9203,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.534570741036045e-05,
+      "loss": 3.9069,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5337337841653658e-05,
+      "loss": 3.9239,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5328951894143138e-05,
+      "loss": 3.9084,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5320565946632618e-05,
+      "loss": 3.9274,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312179999122098e-05,
+      "loss": 3.9205,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303794051611578e-05,
+      "loss": 3.9134,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.529542448290479e-05,
+      "loss": 3.9331,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.528703853539427e-05,
+      "loss": 3.9404,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.527865258788375e-05,
+      "loss": 3.9234,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.527026664037323e-05,
+      "loss": 3.9149,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.526188069286271e-05,
+      "loss": 3.9226,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.525349474535219e-05,
+      "loss": 3.9153,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245108797841667e-05,
+      "loss": 3.9182,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236722850331146e-05,
+      "loss": 3.9338,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228353281624362e-05,
+      "loss": 3.9145,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5219967334113836e-05,
+      "loss": 3.9308,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211581386603315e-05,
+      "loss": 3.9216,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5203195439092795e-05,
+      "loss": 3.9231,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.519482587038601e-05,
+      "loss": 3.9043,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5186456301679224e-05,
+      "loss": 3.9283,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5178070354168704e-05,
+      "loss": 3.9124,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5169684406658184e-05,
+      "loss": 3.9309,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516129845914766e-05,
+      "loss": 3.9225,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.515291251163714e-05,
+      "loss": 3.9292,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5144542942930356e-05,
+      "loss": 3.9083,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5136156995419836e-05,
+      "loss": 3.9303,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.512777104790931e-05,
+      "loss": 3.9217,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511938510039879e-05,
+      "loss": 3.9237,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511099915288827e-05,
+      "loss": 3.9213,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.510261320537775e-05,
+      "loss": 3.9301,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094227257867232e-05,
+      "loss": 3.928,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085841310356712e-05,
+      "loss": 3.9282,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.507747174164992e-05,
+      "loss": 3.9143,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506910217294313e-05,
+      "loss": 3.9172,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506071622543261e-05,
+      "loss": 3.921,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052330277922094e-05,
+      "loss": 3.9117,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5043944330411574e-05,
+      "loss": 3.9233,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035558382901054e-05,
+      "loss": 3.9312,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5027172435390534e-05,
+      "loss": 3.9217,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5018802866683743e-05,
+      "loss": 3.9144,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5010416919173223e-05,
+      "loss": 3.924,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.000246047973633,
+      "eval_runtime": 300.6367,
+      "eval_samples_per_second": 1269.276,
+      "eval_steps_per_second": 39.666,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5002030971662703e-05,
+      "loss": 3.9112,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4993645024152186e-05,
+      "loss": 3.9169,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4985259076641666e-05,
+      "loss": 3.9261,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4976873129131146e-05,
+      "loss": 3.9185,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968487181620626e-05,
+      "loss": 3.9296,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4960101234110106e-05,
+      "loss": 3.9128,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951715286599583e-05,
+      "loss": 3.9219,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4943329339089062e-05,
+      "loss": 3.909,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4934943391578542e-05,
+      "loss": 3.9252,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4926557444068026e-05,
+      "loss": 3.9171,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4918171496557506e-05,
+      "loss": 3.9248,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909785549046986e-05,
+      "loss": 3.9306,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901415980340195e-05,
+      "loss": 3.912,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4893030032829675e-05,
+      "loss": 3.9218,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4884644085319155e-05,
+      "loss": 3.9153,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4876258137808635e-05,
+      "loss": 3.9175,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4867872190298115e-05,
+      "loss": 3.9076,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859486242787595e-05,
+      "loss": 3.9152,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4851100295277075e-05,
+      "loss": 3.913,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842714347766554e-05,
+      "loss": 3.9379,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834328400256034e-05,
+      "loss": 3.9171,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4825942452745514e-05,
+      "loss": 3.9244,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817556505234994e-05,
+      "loss": 3.9304,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4809170557724474e-05,
+      "loss": 3.9306,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800800989017687e-05,
+      "loss": 3.9082,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792415041507167e-05,
+      "loss": 3.9211,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4784045472800376e-05,
+      "loss": 3.9081,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775659525289856e-05,
+      "loss": 3.92,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.476727357777934e-05,
+      "loss": 3.9094,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.475888763026882e-05,
+      "loss": 3.9169,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750501682758296e-05,
+      "loss": 3.9207,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4742115735247776e-05,
+      "loss": 3.9208,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733746166540988e-05,
+      "loss": 3.9138,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725360219030468e-05,
+      "loss": 3.9256,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.471699065032368e-05,
+      "loss": 3.9163,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.470860470281316e-05,
+      "loss": 3.9221,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.470021875530264e-05,
+      "loss": 3.905,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.469183280779212e-05,
+      "loss": 3.9213,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683446860281597e-05,
+      "loss": 3.9064,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4675060912771077e-05,
+      "loss": 3.9128,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666674965260557e-05,
+      "loss": 3.9113,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465828901775004e-05,
+      "loss": 3.9243,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464991944904325e-05,
+      "loss": 3.9197,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464153350153273e-05,
+      "loss": 3.9181,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.463314755402221e-05,
+      "loss": 3.9225,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.462476160651169e-05,
+      "loss": 3.9214,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.461637565900117e-05,
+      "loss": 3.9185,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.460798971149065e-05,
+      "loss": 3.9124,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459962014278386e-05,
+      "loss": 3.9039,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459123419527334e-05,
+      "loss": 3.9264,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.458284824776282e-05,
+      "loss": 3.9131,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4574462300252298e-05,
+      "loss": 3.9077,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.456609273154551e-05,
+      "loss": 3.9103,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557706784034994e-05,
+      "loss": 3.9165,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549320836524474e-05,
+      "loss": 3.9039,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.454093488901395e-05,
+      "loss": 3.9091,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532565320307163e-05,
+      "loss": 3.9145,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4524179372796643e-05,
+      "loss": 3.9132,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4515793425286123e-05,
+      "loss": 3.9253,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4507407477775603e-05,
+      "loss": 3.9036,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4499021530265083e-05,
+      "loss": 3.9072,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490635582754563e-05,
+      "loss": 3.9197,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482249635244043e-05,
+      "loss": 3.9029,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4473880066537252e-05,
+      "loss": 3.9114,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446549411902673e-05,
+      "loss": 3.918,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445710817151621e-05,
+      "loss": 3.9222,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4448722224005695e-05,
+      "loss": 3.9072,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440336276495175e-05,
+      "loss": 3.9004,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4431950328984655e-05,
+      "loss": 3.8974,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.442356438147413e-05,
+      "loss": 3.9148,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.441517843396361e-05,
+      "loss": 3.9286,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4406808865256824e-05,
+      "loss": 3.912,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4398422917746304e-05,
+      "loss": 3.9231,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4390036970235784e-05,
+      "loss": 3.9246,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381651022725264e-05,
+      "loss": 3.9277,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373281454018476e-05,
+      "loss": 3.9049,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364895506507953e-05,
+      "loss": 3.9191,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356509558997433e-05,
+      "loss": 3.9158,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4348123611486913e-05,
+      "loss": 3.9069,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4339737663976396e-05,
+      "loss": 3.9198,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4331351716465876e-05,
+      "loss": 3.915,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4322965768955356e-05,
+      "loss": 3.9156,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4314579821444836e-05,
+      "loss": 3.9177,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4306210252738045e-05,
+      "loss": 3.9127,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297840684031258e-05,
+      "loss": 3.9193,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428947111532447e-05,
+      "loss": 3.9102,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428108516781395e-05,
+      "loss": 3.9152,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4272699220303427e-05,
+      "loss": 3.9116,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264313272792906e-05,
+      "loss": 3.9123,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4255927325282386e-05,
+      "loss": 3.9199,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247541377771866e-05,
+      "loss": 3.9159,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423915543026135e-05,
+      "loss": 3.9081,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423076948275083e-05,
+      "loss": 3.9067,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.422238353524031e-05,
+      "loss": 3.9255,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4213997587729786e-05,
+      "loss": 3.9209,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205628019023e-05,
+      "loss": 3.9157,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.419724207151248e-05,
+      "loss": 3.9056,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418885612400196e-05,
+      "loss": 3.9139,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418047017649144e-05,
+      "loss": 3.909,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.417208422898092e-05,
+      "loss": 3.9239,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.41636982814704e-05,
+      "loss": 3.9223,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.415531233395988e-05,
+      "loss": 3.911,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.414692638644936e-05,
+      "loss": 3.9056,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138556817742567e-05,
+      "loss": 3.9183,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130187249035783e-05,
+      "loss": 3.8882,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.412180130152526e-05,
+      "loss": 3.9137,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.411341535401474e-05,
+      "loss": 3.9044,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.410502940650422e-05,
+      "loss": 3.9118,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.40966434589937e-05,
+      "loss": 3.8988,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.408825751148318e-05,
+      "loss": 3.9153,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079887942776392e-05,
+      "loss": 3.9077,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071501995265872e-05,
+      "loss": 3.9129,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4063116047755352e-05,
+      "loss": 3.9175,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4054730100244832e-05,
+      "loss": 3.9049,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4046344152734312e-05,
+      "loss": 3.9262,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.403795820522379e-05,
+      "loss": 3.9327,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029572257713272e-05,
+      "loss": 3.9159,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021202689006484e-05,
+      "loss": 3.9043,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012816741495964e-05,
+      "loss": 3.9188,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4004447172789174e-05,
+      "loss": 3.9113,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3996061225278653e-05,
+      "loss": 3.9069,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987675277768133e-05,
+      "loss": 3.9261,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3979289330257613e-05,
+      "loss": 3.9064,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3970903382747093e-05,
+      "loss": 3.9213,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962517435236573e-05,
+      "loss": 3.9141,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3954131487726053e-05,
+      "loss": 3.9182,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3945745540215533e-05,
+      "loss": 3.9049,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3937375971508742e-05,
+      "loss": 3.911,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3929006402801958e-05,
+      "loss": 3.9098,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3920620455291438e-05,
+      "loss": 3.9231,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3912234507780915e-05,
+      "loss": 3.9132,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3903848560270395e-05,
+      "loss": 3.9185,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895462612759875e-05,
+      "loss": 3.9023,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3887093044053087e-05,
+      "loss": 3.9228,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3878707096542567e-05,
+      "loss": 3.9145,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387033752783578e-05,
+      "loss": 3.9182,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.386195158032526e-05,
+      "loss": 3.9102,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853565632814736e-05,
+      "loss": 3.9218,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3845179685304216e-05,
+      "loss": 3.9185,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3836793737793696e-05,
+      "loss": 3.9261,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3828407790283176e-05,
+      "loss": 3.9029,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382002184277266e-05,
+      "loss": 3.9144,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381163589526214e-05,
+      "loss": 3.9098,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.380324994775162e-05,
+      "loss": 3.9057,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794864000241096e-05,
+      "loss": 3.9169,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786478052730576e-05,
+      "loss": 3.9201,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3778092105220056e-05,
+      "loss": 3.9172,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3769722536513268e-05,
+      "loss": 3.9018,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376135296780648e-05,
+      "loss": 3.9181,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9965667724609375,
+      "eval_runtime": 306.2951,
+      "eval_samples_per_second": 1245.828,
+      "eval_steps_per_second": 38.933,
+      "step": 1602720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.375296702029596e-05,
+      "loss": 3.9113,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.374458107278544e-05,
+      "loss": 3.9061,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3736195125274917e-05,
+      "loss": 3.9204,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3727809177764397e-05,
+      "loss": 3.9113,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.371942323025388e-05,
+      "loss": 3.917,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.371103728274336e-05,
+      "loss": 3.9091,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.370266771403657e-05,
+      "loss": 3.9121,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.369428176652605e-05,
+      "loss": 3.9028,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.368589581901553e-05,
+      "loss": 3.9154,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.367750987150501e-05,
+      "loss": 3.9122,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.366912392399449e-05,
+      "loss": 3.9145,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3660737976483973e-05,
+      "loss": 3.9203,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3652368407777182e-05,
+      "loss": 3.9064,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.364398246026666e-05,
+      "loss": 3.9151,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.363559651275614e-05,
+      "loss": 3.9107,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.362721056524562e-05,
+      "loss": 3.9079,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3618840996538834e-05,
+      "loss": 3.8987,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3610455049028314e-05,
+      "loss": 3.9076,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3602069101517794e-05,
+      "loss": 3.904,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3593683154007274e-05,
+      "loss": 3.9264,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.358529720649675e-05,
+      "loss": 3.9128,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.357691125898623e-05,
+      "loss": 3.9158,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.356852531147571e-05,
+      "loss": 3.9237,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.356013936396519e-05,
+      "loss": 3.9219,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3551769795258403e-05,
+      "loss": 3.8981,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3543383847747883e-05,
+      "loss": 3.9139,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3534997900237363e-05,
+      "loss": 3.9054,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3526628331530572e-05,
+      "loss": 3.9095,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3518242384020052e-05,
+      "loss": 3.9032,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3509856436509535e-05,
+      "loss": 3.9079,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3501470488999015e-05,
+      "loss": 3.9166,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3493084541488495e-05,
+      "loss": 3.9105,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3484698593977975e-05,
+      "loss": 3.9105,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3476312646467455e-05,
+      "loss": 3.9145,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3467959456564397e-05,
+      "loss": 3.9126,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3459573509053877e-05,
+      "loss": 3.9127,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3451187561543357e-05,
+      "loss": 3.9023,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3442801614032836e-05,
+      "loss": 3.9055,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3434415666522316e-05,
+      "loss": 3.9038,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3426029719011796e-05,
+      "loss": 3.9035,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3417643771501276e-05,
+      "loss": 3.9048,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3409257823990753e-05,
+      "loss": 3.9159,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3400871876480236e-05,
+      "loss": 3.9137,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3392485928969716e-05,
+      "loss": 3.9095,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3384116360262925e-05,
+      "loss": 3.9197,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3375730412752405e-05,
+      "loss": 3.9099,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3367344465241885e-05,
+      "loss": 3.9145,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3358958517731365e-05,
+      "loss": 3.9051,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3350588949024578e-05,
+      "loss": 3.8932,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3342203001514058e-05,
+      "loss": 3.9209,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3333817054003538e-05,
+      "loss": 3.9041,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3325431106493018e-05,
+      "loss": 3.9049,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3317045158982497e-05,
+      "loss": 3.9036,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3308659211471977e-05,
+      "loss": 3.9036,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.330028964276519e-05,
+      "loss": 3.8973,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.329190369525467e-05,
+      "loss": 3.9017,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.328351774774415e-05,
+      "loss": 3.9075,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.327513180023363e-05,
+      "loss": 3.9063,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3266745852723106e-05,
+      "loss": 3.9167,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3258359905212586e-05,
+      "loss": 3.8963,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3249973957702066e-05,
+      "loss": 3.9017,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3241588010191546e-05,
+      "loss": 3.9142,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.323321844148476e-05,
+      "loss": 3.8964,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.322483249397424e-05,
+      "loss": 3.8997,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.321644654646372e-05,
+      "loss": 3.9072,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.32080605989532e-05,
+      "loss": 3.9161,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3199691030246408e-05,
+      "loss": 3.8988,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.319130508273589e-05,
+      "loss": 3.8982,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.318291913522537e-05,
+      "loss": 3.8878,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.317453318771485e-05,
+      "loss": 3.9076,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.316614724020433e-05,
+      "loss": 3.9197,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.315777767149754e-05,
+      "loss": 3.9071,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.314939172398702e-05,
+      "loss": 3.9107,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3141022155280232e-05,
+      "loss": 3.9162,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3132636207769712e-05,
+      "loss": 3.9234,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3124250260259192e-05,
+      "loss": 3.8967,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3115864312748672e-05,
+      "loss": 3.912,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.310749474404188e-05,
+      "loss": 3.9085,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.309910879653136e-05,
+      "loss": 3.8954,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3090722849020845e-05,
+      "loss": 3.9118,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3082336901510325e-05,
+      "loss": 3.9097,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3073950953999805e-05,
+      "loss": 3.9074,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3065581385293014e-05,
+      "loss": 3.913,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3057195437782494e-05,
+      "loss": 3.9052,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3048809490271974e-05,
+      "loss": 3.9032,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3040423542761454e-05,
+      "loss": 3.9057,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3032037595250937e-05,
+      "loss": 3.9113,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3023651647740413e-05,
+      "loss": 3.9011,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3015282079033626e-05,
+      "loss": 3.9042,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3006896131523106e-05,
+      "loss": 3.9118,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2998510184012586e-05,
+      "loss": 3.9094,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2990124236502066e-05,
+      "loss": 3.9017,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2981738288991546e-05,
+      "loss": 3.9009,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2973352341481026e-05,
+      "loss": 3.9201,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2964966393970506e-05,
+      "loss": 3.9124,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2956580446459986e-05,
+      "loss": 3.9123,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2948210877753195e-05,
+      "loss": 3.897,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2939824930242675e-05,
+      "loss": 3.906,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2931438982732155e-05,
+      "loss": 3.9042,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2923053035221638e-05,
+      "loss": 3.9141,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2914667087711118e-05,
+      "loss": 3.9164,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2906297519004327e-05,
+      "loss": 3.9035,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2897911571493807e-05,
+      "loss": 3.8983,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2889525623983287e-05,
+      "loss": 3.911,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.28811560552765e-05,
+      "loss": 3.8822,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.287277010776598e-05,
+      "loss": 3.9041,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.286438416025546e-05,
+      "loss": 3.8916,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.285599821274494e-05,
+      "loss": 3.909,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2847612265234416e-05,
+      "loss": 3.8939,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.283924269652763e-05,
+      "loss": 3.907,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2830856749017108e-05,
+      "loss": 3.8979,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.282247080150659e-05,
+      "loss": 3.9046,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2814084853996068e-05,
+      "loss": 3.9096,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2805698906485548e-05,
+      "loss": 3.9029,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2797312958975028e-05,
+      "loss": 3.9168,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2788927011464508e-05,
+      "loss": 3.9225,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2780541063953988e-05,
+      "loss": 3.9094,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.27721714952472e-05,
+      "loss": 3.8999,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.276378554773668e-05,
+      "loss": 3.9045,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.275539960022616e-05,
+      "loss": 3.9066,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.274701365271564e-05,
+      "loss": 3.8944,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.273864408400885e-05,
+      "loss": 3.9234,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.273025813649833e-05,
+      "loss": 3.8956,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2721888567791542e-05,
+      "loss": 3.9158,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2713502620281022e-05,
+      "loss": 3.9064,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2705116672770502e-05,
+      "loss": 3.9068,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2696730725259982e-05,
+      "loss": 3.8963,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.268836115655319e-05,
+      "loss": 3.9077,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2679975209042674e-05,
+      "loss": 3.9019,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2671589261532154e-05,
+      "loss": 3.9138,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2663203314021634e-05,
+      "loss": 3.9069,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2654817366511114e-05,
+      "loss": 3.9111,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2646431419000594e-05,
+      "loss": 3.8945,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.263804547149007e-05,
+      "loss": 3.9166,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.262965952397955e-05,
+      "loss": 3.9056,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2621289955272763e-05,
+      "loss": 3.9112,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2612904007762246e-05,
+      "loss": 3.9061,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2604518060251723e-05,
+      "loss": 3.9128,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2596132112741203e-05,
+      "loss": 3.91,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2587762544034415e-05,
+      "loss": 3.9225,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2579376596523892e-05,
+      "loss": 3.894,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2570990649013375e-05,
+      "loss": 3.9047,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2562604701502855e-05,
+      "loss": 3.9032,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2554218753992335e-05,
+      "loss": 3.9027,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2545849185285544e-05,
+      "loss": 3.9034,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2537463237775024e-05,
+      "loss": 3.9118,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2529077290264504e-05,
+      "loss": 3.9135,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2520691342753984e-05,
+      "loss": 3.8959,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2512321774047197e-05,
+      "loss": 3.9132,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.993469476699829,
+      "eval_runtime": 304.4944,
+      "eval_samples_per_second": 1253.196,
+      "eval_steps_per_second": 39.163,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2503935826536677e-05,
+      "loss": 3.9057,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2495549879026157e-05,
+      "loss": 3.8957,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2487163931515637e-05,
+      "loss": 3.916,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2478777984005117e-05,
+      "loss": 3.9039,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2470392036494596e-05,
+      "loss": 3.9119,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2462006088984076e-05,
+      "loss": 3.9019,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2453620141473556e-05,
+      "loss": 3.9023,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2445234193963036e-05,
+      "loss": 3.8976,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2436848246452516e-05,
+      "loss": 3.9084,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2428462298941996e-05,
+      "loss": 3.9025,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2420076351431476e-05,
+      "loss": 3.91,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2411690403920956e-05,
+      "loss": 3.9137,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.240332083521417e-05,
+      "loss": 3.9033,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239493488770365e-05,
+      "loss": 3.9043,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238654894019313e-05,
+      "loss": 3.9078,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.237816299268261e-05,
+      "loss": 3.8996,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2369777045172085e-05,
+      "loss": 3.8921,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2361391097661565e-05,
+      "loss": 3.9016,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2353005150151045e-05,
+      "loss": 3.8996,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2344619202640525e-05,
+      "loss": 3.9149,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2336249633933737e-05,
+      "loss": 3.9085,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2327863686423217e-05,
+      "loss": 3.9097,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2319477738912697e-05,
+      "loss": 3.9132,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2311091791402177e-05,
+      "loss": 3.914,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2302705843891657e-05,
+      "loss": 3.8962,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229433627518487e-05,
+      "loss": 3.9084,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.228595032767435e-05,
+      "loss": 3.8924,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227756438016383e-05,
+      "loss": 3.9026,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.226917843265331e-05,
+      "loss": 3.9027,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.226080886394652e-05,
+      "loss": 3.899,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252422916436e-05,
+      "loss": 3.9084,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.224403696892548e-05,
+      "loss": 3.905,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235651021414962e-05,
+      "loss": 3.9052,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.222726507390444e-05,
+      "loss": 3.9078,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.221889550519765e-05,
+      "loss": 3.9026,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.221050955768713e-05,
+      "loss": 3.9075,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.220212361017661e-05,
+      "loss": 3.8964,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.219373766266609e-05,
+      "loss": 3.8987,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.218535171515557e-05,
+      "loss": 3.8956,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2176982146448783e-05,
+      "loss": 3.8979,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.216859619893826e-05,
+      "loss": 3.8977,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.216021025142774e-05,
+      "loss": 3.9083,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.215182430391722e-05,
+      "loss": 3.9047,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.21434383564067e-05,
+      "loss": 3.9042,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.213505240889618e-05,
+      "loss": 3.9134,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126682840189392e-05,
+      "loss": 3.905,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118296892678872e-05,
+      "loss": 3.9083,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109910945168352e-05,
+      "loss": 3.8944,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101524997657832e-05,
+      "loss": 3.8852,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2093139050147312e-05,
+      "loss": 3.9163,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084753102636792e-05,
+      "loss": 3.8988,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076383533930004e-05,
+      "loss": 3.8935,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2067997586419484e-05,
+      "loss": 3.9017,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2059611638908964e-05,
+      "loss": 3.8942,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205122569139844e-05,
+      "loss": 3.8958,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204283974388792e-05,
+      "loss": 3.8899,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.20344537963774e-05,
+      "loss": 3.9006,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2026067848866884e-05,
+      "loss": 3.8997,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017681901356364e-05,
+      "loss": 3.9107,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2009312332649573e-05,
+      "loss": 3.8901,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000926385139053e-05,
+      "loss": 3.8904,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1992556816432266e-05,
+      "loss": 3.9127,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1984170868921746e-05,
+      "loss": 3.8853,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1975784921411226e-05,
+      "loss": 3.8936,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1967398973900706e-05,
+      "loss": 3.9031,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1959013026390185e-05,
+      "loss": 3.9125,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1950627078879665e-05,
+      "loss": 3.8915,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1942241131369145e-05,
+      "loss": 3.8921,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1933855183858622e-05,
+      "loss": 3.8812,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1925469236348102e-05,
+      "loss": 3.8985,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1917099667641318e-05,
+      "loss": 3.9149,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1908730098934527e-05,
+      "loss": 3.9015,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1900344151424007e-05,
+      "loss": 3.9018,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1891958203913487e-05,
+      "loss": 3.9104,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1883572256402967e-05,
+      "loss": 3.9154,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187520268769618e-05,
+      "loss": 3.8934,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186681674018566e-05,
+      "loss": 3.9038,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1858447171478868e-05,
+      "loss": 3.9043,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1850061223968348e-05,
+      "loss": 3.8895,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1841675276457828e-05,
+      "loss": 3.9043,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1833289328947308e-05,
+      "loss": 3.9037,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1824903381436788e-05,
+      "loss": 3.9029,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.181653381273e-05,
+      "loss": 3.9054,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.180814786521948e-05,
+      "loss": 3.902,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.179976191770896e-05,
+      "loss": 3.8948,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.179137597019844e-05,
+      "loss": 3.9021,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.178299002268792e-05,
+      "loss": 3.9015,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774620453981133e-05,
+      "loss": 3.8978,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766234506470613e-05,
+      "loss": 3.9001,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757848558960093e-05,
+      "loss": 3.9,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.174946261144957e-05,
+      "loss": 3.9058,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.174107666393905e-05,
+      "loss": 3.8972,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.173269071642853e-05,
+      "loss": 3.8889,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.172430476891801e-05,
+      "loss": 3.9143,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715918821407493e-05,
+      "loss": 3.9036,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707532873896973e-05,
+      "loss": 3.906,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1699146926386453e-05,
+      "loss": 3.8901,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169076097887593e-05,
+      "loss": 3.9026,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.168237503136541e-05,
+      "loss": 3.8959,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1674021841462354e-05,
+      "loss": 3.9048,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665635893951834e-05,
+      "loss": 3.9144,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657249946441314e-05,
+      "loss": 3.8974,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648863998930794e-05,
+      "loss": 3.8953,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1640478051420274e-05,
+      "loss": 3.9008,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.163209210390975e-05,
+      "loss": 3.8764,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.162370615639923e-05,
+      "loss": 3.9027,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615336587692446e-05,
+      "loss": 3.8817,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606950640181926e-05,
+      "loss": 3.9022,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598564692671403e-05,
+      "loss": 3.8888,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1590178745160883e-05,
+      "loss": 3.8993,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581792797650363e-05,
+      "loss": 3.8937,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1573406850139843e-05,
+      "loss": 3.8962,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1565020902629323e-05,
+      "loss": 3.9081,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1556651333922535e-05,
+      "loss": 3.8945,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1548265386412015e-05,
+      "loss": 3.9104,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1539879438901495e-05,
+      "loss": 3.9154,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1531493491390975e-05,
+      "loss": 3.9023,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1523107543880455e-05,
+      "loss": 3.9002,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.151472159636993e-05,
+      "loss": 3.8935,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1506352027663147e-05,
+      "loss": 3.9011,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1497966080152627e-05,
+      "loss": 3.8971,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489580132642107e-05,
+      "loss": 3.9152,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1481194185131584e-05,
+      "loss": 3.8869,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1472824616424796e-05,
+      "loss": 3.909,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1464438668914276e-05,
+      "loss": 3.9039,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1456052721403756e-05,
+      "loss": 3.9005,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1447666773893236e-05,
+      "loss": 3.8891,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1439280826382716e-05,
+      "loss": 3.9034,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1430894878872196e-05,
+      "loss": 3.8932,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1422508931361676e-05,
+      "loss": 3.9123,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1414122983851156e-05,
+      "loss": 3.8982,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1405753415144365e-05,
+      "loss": 3.9062,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.139736746763385e-05,
+      "loss": 3.8879,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138898152012333e-05,
+      "loss": 3.9119,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138059557261281e-05,
+      "loss": 3.9004,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1372226003906017e-05,
+      "loss": 3.9063,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1363840056395497e-05,
+      "loss": 3.8998,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.135547048768871e-05,
+      "loss": 3.9013,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.134708454017819e-05,
+      "loss": 3.9061,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133869859266767e-05,
+      "loss": 3.9203,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133031264515715e-05,
+      "loss": 3.8867,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.132192669764663e-05,
+      "loss": 3.8999,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.131354075013611e-05,
+      "loss": 3.8968,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1305154802625586e-05,
+      "loss": 3.8991,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.129676885511507e-05,
+      "loss": 3.8932,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288399286408282e-05,
+      "loss": 3.9035,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.128002971770149e-05,
+      "loss": 3.9057,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127164377019097e-05,
+      "loss": 3.8906,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.126325782268045e-05,
+      "loss": 3.9076,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.991149425506592,
+      "eval_runtime": 310.2187,
+      "eval_samples_per_second": 1230.071,
+      "eval_steps_per_second": 38.441,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.125487187516993e-05,
+      "loss": 3.8967,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.124648592765941e-05,
+      "loss": 3.8911,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1238116358952623e-05,
+      "loss": 3.9077,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1229730411442103e-05,
+      "loss": 3.8943,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1221344463931583e-05,
+      "loss": 3.9043,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.121295851642106e-05,
+      "loss": 3.8961,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.120457256891054e-05,
+      "loss": 3.8989,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.119618662140002e-05,
+      "loss": 3.8918,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1187800673889503e-05,
+      "loss": 3.9035,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1179431105182712e-05,
+      "loss": 3.8939,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1171045157672192e-05,
+      "loss": 3.9052,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1162659210161672e-05,
+      "loss": 3.9057,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115428964145488e-05,
+      "loss": 3.9003,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1145903693944365e-05,
+      "loss": 3.8993,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1137517746433845e-05,
+      "loss": 3.8971,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1129131798923325e-05,
+      "loss": 3.8907,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1120745851412805e-05,
+      "loss": 3.889,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1112359903902284e-05,
+      "loss": 3.8948,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1103973956391764e-05,
+      "loss": 3.8973,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.109558800888124e-05,
+      "loss": 3.9027,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1087202061370724e-05,
+      "loss": 3.9064,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1078816113860204e-05,
+      "loss": 3.9027,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1070430166349684e-05,
+      "loss": 3.9048,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1062060597642893e-05,
+      "loss": 3.9081,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1053674650132373e-05,
+      "loss": 3.8868,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1045288702621853e-05,
+      "loss": 3.9037,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1036902755111333e-05,
+      "loss": 3.891,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1028533186404546e-05,
+      "loss": 3.8955,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1020147238894026e-05,
+      "loss": 3.8927,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1011761291383506e-05,
+      "loss": 3.8985,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1003375343872986e-05,
+      "loss": 3.8992,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0994989396362466e-05,
+      "loss": 3.9015,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0986603448851945e-05,
+      "loss": 3.8983,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0978217501341425e-05,
+      "loss": 3.9027,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0969847932634638e-05,
+      "loss": 3.8923,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0961461985124118e-05,
+      "loss": 3.9037,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0953076037613594e-05,
+      "loss": 3.8993,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0944690090103074e-05,
+      "loss": 3.8858,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0936304142592554e-05,
+      "loss": 3.8888,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0927918195082034e-05,
+      "loss": 3.8902,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0919532247571518e-05,
+      "loss": 3.8976,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0911146300060998e-05,
+      "loss": 3.9008,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0902776731354207e-05,
+      "loss": 3.9004,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894390783843687e-05,
+      "loss": 3.8959,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0886004836333167e-05,
+      "loss": 3.9091,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877618888822647e-05,
+      "loss": 3.8995,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0869232941312127e-05,
+      "loss": 3.9051,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.086086337260534e-05,
+      "loss": 3.8921,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.085247742509482e-05,
+      "loss": 3.8733,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.08440914775843e-05,
+      "loss": 3.9118,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0835705530073776e-05,
+      "loss": 3.8947,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0827335961366988e-05,
+      "loss": 3.8922,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.081895001385647e-05,
+      "loss": 3.8928,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.081056406634595e-05,
+      "loss": 3.8886,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0802178118835428e-05,
+      "loss": 3.8907,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.079380855012864e-05,
+      "loss": 3.8825,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.078542260261812e-05,
+      "loss": 3.8956,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.07770366551076e-05,
+      "loss": 3.8906,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076865070759708e-05,
+      "loss": 3.908,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076026476008656e-05,
+      "loss": 3.8857,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.075187881257604e-05,
+      "loss": 3.8855,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.074349286506552e-05,
+      "loss": 3.9046,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0735106917555e-05,
+      "loss": 3.8775,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.072672097004448e-05,
+      "loss": 3.8872,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.071835140133769e-05,
+      "loss": 3.9002,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0709965453827172e-05,
+      "loss": 3.9047,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.070159588512038e-05,
+      "loss": 3.8875,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.069320993760986e-05,
+      "loss": 3.8868,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.068482399009934e-05,
+      "loss": 3.878,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.067643804258882e-05,
+      "loss": 3.8913,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.06680520950783e-05,
+      "loss": 3.9086,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659682526371514e-05,
+      "loss": 3.8942,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651296578860994e-05,
+      "loss": 3.8965,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0642910631350474e-05,
+      "loss": 3.9062,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0634524683839954e-05,
+      "loss": 3.9099,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0626155115133163e-05,
+      "loss": 3.8853,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0617769167622643e-05,
+      "loss": 3.9012,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0609383220112126e-05,
+      "loss": 3.8906,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0600997272601606e-05,
+      "loss": 3.8914,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0592611325091083e-05,
+      "loss": 3.8933,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0584225377580563e-05,
+      "loss": 3.9,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0575855808873775e-05,
+      "loss": 3.8954,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0567469861363255e-05,
+      "loss": 3.9028,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0559083913852735e-05,
+      "loss": 3.8944,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0550697966342215e-05,
+      "loss": 3.8854,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0542312018831695e-05,
+      "loss": 3.8969,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0533926071321175e-05,
+      "loss": 3.8946,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0525540123810655e-05,
+      "loss": 3.8914,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0517154176300135e-05,
+      "loss": 3.892,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0508784607593344e-05,
+      "loss": 3.8955,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0500398660082827e-05,
+      "loss": 3.8992,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0492012712572307e-05,
+      "loss": 3.8936,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0483626765061787e-05,
+      "loss": 3.886,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0475240817551264e-05,
+      "loss": 3.9052,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0466854870040744e-05,
+      "loss": 3.8992,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0458485301333956e-05,
+      "loss": 3.8994,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0450099353823436e-05,
+      "loss": 3.8857,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0441713406312916e-05,
+      "loss": 3.8947,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0433327458802396e-05,
+      "loss": 3.8896,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.042495789009561e-05,
+      "loss": 3.8976,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416571942585085e-05,
+      "loss": 3.9049,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0408185995074565e-05,
+      "loss": 3.895,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0399800047564045e-05,
+      "loss": 3.8916,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0391414100053528e-05,
+      "loss": 3.889,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0383028152543008e-05,
+      "loss": 3.878,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0374642205032488e-05,
+      "loss": 3.8928,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0366256257521968e-05,
+      "loss": 3.8782,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0357870310011445e-05,
+      "loss": 3.897,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0349500741304657e-05,
+      "loss": 3.8829,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0341114793794137e-05,
+      "loss": 3.8919,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0332728846283617e-05,
+      "loss": 3.887,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0324342898773097e-05,
+      "loss": 3.8959,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0315956951262577e-05,
+      "loss": 3.8994,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.030758738255579e-05,
+      "loss": 3.8882,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0299201435045266e-05,
+      "loss": 3.9059,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.029081548753475e-05,
+      "loss": 3.9046,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.028242954002423e-05,
+      "loss": 3.8942,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.027404359251371e-05,
+      "loss": 3.897,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.026565764500319e-05,
+      "loss": 3.8849,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025727169749267e-05,
+      "loss": 3.8943,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.024888574998215e-05,
+      "loss": 3.8946,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0240516181275358e-05,
+      "loss": 3.9018,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.023213023376484e-05,
+      "loss": 3.8839,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.022374428625432e-05,
+      "loss": 3.9019,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0215358338743798e-05,
+      "loss": 3.9005,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0206972391233278e-05,
+      "loss": 3.8948,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0198586443722758e-05,
+      "loss": 3.881,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0190200496212238e-05,
+      "loss": 3.8976,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0181814548701718e-05,
+      "loss": 3.8885,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.017344497999493e-05,
+      "loss": 3.9058,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.016505903248441e-05,
+      "loss": 3.8978,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015667308497389e-05,
+      "loss": 3.8985,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.014828713746337e-05,
+      "loss": 3.8811,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013991756875658e-05,
+      "loss": 3.9108,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013153162124606e-05,
+      "loss": 3.8908,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0123145673735543e-05,
+      "loss": 3.899,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0114759726225023e-05,
+      "loss": 3.8949,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0106373778714503e-05,
+      "loss": 3.8945,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0098004210007712e-05,
+      "loss": 3.8997,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008961826249719e-05,
+      "loss": 3.9104,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008123231498667e-05,
+      "loss": 3.882,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.007284636747615e-05,
+      "loss": 3.8974,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.006446041996563e-05,
+      "loss": 3.8925,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0056107230062573e-05,
+      "loss": 3.8901,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0047721282552053e-05,
+      "loss": 3.8899,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0039335335041533e-05,
+      "loss": 3.8988,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0030949387531013e-05,
+      "loss": 3.9,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0022563440020496e-05,
+      "loss": 3.8847,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0014177492509976e-05,
+      "loss": 3.9016,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9886837005615234,
+      "eval_runtime": 337.1629,
+      "eval_samples_per_second": 1131.771,
+      "eval_steps_per_second": 35.369,
+      "step": 1831680
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 7.620243884277248e+17,
+  "trial_name": null,
+  "trial_params": null
+}