diff --git "a/checkpoint-1984320/trainer_state.json" "b/checkpoint-1984320/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1984320/trainer_state.json"
@@ -0,0 +1,23483 @@
+{
+  "best_metric": 3.893688440322876,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/transformer/2/checkpoints/checkpoint-839520",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1984320,
+  "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": 11.0111,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8304,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1934,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9809,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8098,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7038,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6068,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5374,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4546,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4007,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3464,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3083,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2703,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2131,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1814,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1288,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1136,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0782,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0411,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0188,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0187,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9727,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9591,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9324,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9072,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8975,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8631,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.859,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8412,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8142,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8068,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 4.7844,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 4.7855,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7661,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7484,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.745,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7463,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7079,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.7015,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.695,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 4.6749,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 4.6688,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 4.6714,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6598,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6449,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 4.6366,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 4.6267,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 4.6274,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5957,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5957,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6022,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5928,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5793,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5615,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 4.5695,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 4.5506,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 4.5676,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 4.5352,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5489,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 4.5302,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 4.5188,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 4.5224,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 4.5084,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 4.5034,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 4.4952,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 4.505,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 4.4809,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 4.4763,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 4.4645,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4629,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4826,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 4.4715,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 4.4523,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 4.4521,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 4.4569,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 4.4513,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 4.4344,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 4.439,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 4.4291,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 4.4267,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 4.4215,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 4.4195,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4217,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4137,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 4.393,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 4.4039,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 4.4025,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 4.3979,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 4.3908,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 4.3806,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 4.3867,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 4.3744,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 4.3819,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 4.3697,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 4.3645,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 4.3834,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.3487,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.3473,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.3556,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.3541,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.3581,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.3437,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.3372,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.3352,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.3373,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.3218,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.3192,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.3317,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.3218,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.3133,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.3101,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.3068,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.3177,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.3039,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.3018,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.3053,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.3055,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.2952,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.2981,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.2947,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.2937,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.2765,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.2885,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.2835,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.2888,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.271,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.2794,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.2694,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.2729,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.2516,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.2786,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.2767,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.2601,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.2503,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766433672724e-05,
+      "loss": 4.2709,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886825741976188e-05,
+      "loss": 4.2564,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.2585,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.2507,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.2591,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.2503,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.2484,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.235,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.2468,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.2427,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.2355,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.2551,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.2296,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.2371,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.2402,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.2283,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.242136001586914,
+      "eval_runtime": 304.3659,
+      "eval_samples_per_second": 1253.725,
+      "eval_steps_per_second": 39.18,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.2164,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.2205,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.2306,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.2205,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.2211,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87006039847664e-05,
+      "loss": 4.2057,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.2073,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.2047,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.2127,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.1982,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.2066,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.21,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.1932,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.191,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.1788,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.1883,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.1836,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.1786,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.1792,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.2071,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.1769,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.178,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558092013498754e-05,
+      "loss": 4.1857,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.1767,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.1722,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.1653,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.1699,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.1672,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.1534,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.1661,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.1531,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.1674,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.1631,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.1475,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.1639,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.169,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.1409,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.1518,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.1438,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.1371,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.1428,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.151,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.1471,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.1387,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.1449,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.1333,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.1414,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.1206,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.1247,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.1389,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.1385,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.1259,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.1163,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.1294,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.1106,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.1372,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.1185,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.1334,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.1184,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.1088,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.1139,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.1173,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.1048,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.103,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.1186,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.1021,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.0946,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.0924,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.0935,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.1131,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.111,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814737713112805e-05,
+      "loss": 4.0958,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.0989,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.1028,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.1,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.0913,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.0992,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.0928,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.0971,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808032230865135e-05,
+      "loss": 4.0878,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.0914,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.0925,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.0914,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.0731,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.0883,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.084,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.0861,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.0826,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.0697,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.083,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.0743,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79797400749363e-05,
+      "loss": 4.0839,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.0678,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796296817991526e-05,
+      "loss": 4.0737,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.085,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.0719,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.0545,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929440768676916e-05,
+      "loss": 4.0684,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921054821166396e-05,
+      "loss": 4.0729,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.074,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904282926145356e-05,
+      "loss": 4.0667,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895896978634836e-05,
+      "loss": 4.0613,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887527409928045e-05,
+      "loss": 4.0587,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879141462417525e-05,
+      "loss": 4.0689,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870755514907005e-05,
+      "loss": 4.0494,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862385946200214e-05,
+      "loss": 4.0546,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853999998689694e-05,
+      "loss": 4.0633,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784561405117918e-05,
+      "loss": 4.0634,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783722810366866e-05,
+      "loss": 4.0416,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782884215615814e-05,
+      "loss": 4.0557,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.0444,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.0614,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.0484,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.043,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.0614,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777854284989875e-05,
+      "loss": 4.0539,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.0473,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.0521,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.048,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.0502,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.0333,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.0525,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.0402,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.05,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703102079911534e-05,
+      "loss": 4.0403,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694716132401014e-05,
+      "loss": 4.0418,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686330184890494e-05,
+      "loss": 4.036,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767794423737997e-05,
+      "loss": 4.0399,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766957466867318e-05,
+      "loss": 4.0225,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766118872116266e-05,
+      "loss": 4.0514,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652802773652136e-05,
+      "loss": 4.0432,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644416826141616e-05,
+      "loss": 4.0364,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763604725743483e-05,
+      "loss": 4.0277,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762766130992431e-05,
+      "loss": 4.0449,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761927536241379e-05,
+      "loss": 4.04,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761088941490327e-05,
+      "loss": 4.0326,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760251984619649e-05,
+      "loss": 4.0287,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759413389868596e-05,
+      "loss": 4.0383,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758574795117544e-05,
+      "loss": 4.0326,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757736200366492e-05,
+      "loss": 4.0423,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568992434958137e-05,
+      "loss": 4.0168,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756060648744761e-05,
+      "loss": 4.0328,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755222053993709e-05,
+      "loss": 4.0293,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754383459242657e-05,
+      "loss": 4.0217,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535465023719786e-05,
+      "loss": 4.0447,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527079076209266e-05,
+      "loss": 4.0225,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518693128698745e-05,
+      "loss": 4.029,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510307181188225e-05,
+      "loss": 4.0274,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.0271,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.07383918762207,
+      "eval_runtime": 304.4679,
+      "eval_samples_per_second": 1253.304,
+      "eval_steps_per_second": 39.167,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493535286167185e-05,
+      "loss": 4.0117,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485149338656665e-05,
+      "loss": 4.0201,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476763391146145e-05,
+      "loss": 4.0274,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468377443635625e-05,
+      "loss": 4.0218,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459991496125105e-05,
+      "loss": 4.0283,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451605548614585e-05,
+      "loss": 4.0047,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443219601104065e-05,
+      "loss": 4.012,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434833653593545e-05,
+      "loss": 4.0094,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426447706083025e-05,
+      "loss": 4.02,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418061758572505e-05,
+      "loss": 4.0096,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740969218986572e-05,
+      "loss": 4.0155,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74013062423552e-05,
+      "loss": 4.0195,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739293667364841e-05,
+      "loss": 4.0025,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738455072613789e-05,
+      "loss": 4.0061,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737616477862737e-05,
+      "loss": 3.9889,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736777883111685e-05,
+      "loss": 4.0012,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735939288360632e-05,
+      "loss": 4.0009,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73510069360958e-05,
+      "loss": 3.9965,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734262098858528e-05,
+      "loss": 3.9973,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733423504107476e-05,
+      "loss": 4.0216,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732584909356424e-05,
+      "loss": 3.9989,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731746314605372e-05,
+      "loss": 3.9976,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73090771985432e-05,
+      "loss": 4.005,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730070762983642e-05,
+      "loss": 4.002,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72923216823259e-05,
+      "loss": 3.9925,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728393573481538e-05,
+      "loss": 3.9884,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727554978730486e-05,
+      "loss": 3.9992,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726718021859807e-05,
+      "loss": 3.9885,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725879427108755e-05,
+      "loss": 3.9824,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725040832357703e-05,
+      "loss": 3.9946,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724202237606651e-05,
+      "loss": 3.9858,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233652807359716e-05,
+      "loss": 3.9962,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225266859849196e-05,
+      "loss": 3.9918,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216880912338676e-05,
+      "loss": 3.9797,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208494964828156e-05,
+      "loss": 3.9942,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720012539612137e-05,
+      "loss": 4.0048,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719173944861085e-05,
+      "loss": 3.9724,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718335350110033e-05,
+      "loss": 3.9881,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717496755358981e-05,
+      "loss": 3.9769,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716659798488302e-05,
+      "loss": 3.9782,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71582120373725e-05,
+      "loss": 3.9769,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714982608986198e-05,
+      "loss": 3.9925,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714144014235146e-05,
+      "loss": 3.9858,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713307057364467e-05,
+      "loss": 3.9793,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712468462613415e-05,
+      "loss": 3.9843,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711629867862363e-05,
+      "loss": 3.9729,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710791273111311e-05,
+      "loss": 3.9861,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099543162406326e-05,
+      "loss": 3.9657,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091157214895806e-05,
+      "loss": 3.9698,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082771267385286e-05,
+      "loss": 3.982,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074385319874766e-05,
+      "loss": 3.9796,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7066015751167975e-05,
+      "loss": 3.9719,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057629803657455e-05,
+      "loss": 3.9668,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049243856146935e-05,
+      "loss": 3.976,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040857908636415e-05,
+      "loss": 3.9563,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032488339929624e-05,
+      "loss": 3.9814,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024102392419104e-05,
+      "loss": 3.9718,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015716444908584e-05,
+      "loss": 3.9828,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007330497398064e-05,
+      "loss": 3.974,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699896092869128e-05,
+      "loss": 3.9605,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699057498118076e-05,
+      "loss": 3.9639,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698218903367024e-05,
+      "loss": 3.9705,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697380308615972e-05,
+      "loss": 3.9572,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696543351745293e-05,
+      "loss": 3.9581,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695704756994241e-05,
+      "loss": 3.9704,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694866162243189e-05,
+      "loss": 3.9603,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694027567492137e-05,
+      "loss": 3.9496,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693190610621458e-05,
+      "loss": 3.9512,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692352015870406e-05,
+      "loss": 3.9499,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691513421119354e-05,
+      "loss": 3.9706,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690674826368302e-05,
+      "loss": 3.9663,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689837869497623e-05,
+      "loss": 3.9568,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688999274746571e-05,
+      "loss": 3.9621,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688160679995519e-05,
+      "loss": 3.9582,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687322085244467e-05,
+      "loss": 3.9621,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686485128373788e-05,
+      "loss": 3.9527,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685648171503109e-05,
+      "loss": 3.9625,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684809576752057e-05,
+      "loss": 3.9574,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683970982001005e-05,
+      "loss": 3.9559,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683132387249953e-05,
+      "loss": 3.9513,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682293792498901e-05,
+      "loss": 3.9577,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681455197747849e-05,
+      "loss": 3.9568,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680616602996797e-05,
+      "loss": 3.9568,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679779646126119e-05,
+      "loss": 3.9437,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678941051375067e-05,
+      "loss": 3.9516,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678102456624015e-05,
+      "loss": 3.9498,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677263861872963e-05,
+      "loss": 3.9508,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764269050022836e-05,
+      "loss": 3.9521,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755883102512316e-05,
+      "loss": 3.9366,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747497155001796e-05,
+      "loss": 3.9485,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739111207491276e-05,
+      "loss": 3.9459,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730741638784485e-05,
+      "loss": 3.9517,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722355691273965e-05,
+      "loss": 3.9352,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713969743763445e-05,
+      "loss": 3.9445,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705583796252925e-05,
+      "loss": 3.9573,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697197848742405e-05,
+      "loss": 3.9435,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688811901231885e-05,
+      "loss": 3.9273,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680425953721365e-05,
+      "loss": 3.9407,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672040006210844e-05,
+      "loss": 3.9464,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663670437504054e-05,
+      "loss": 3.9479,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655284489993534e-05,
+      "loss": 3.9421,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646898542483013e-05,
+      "loss": 3.9321,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638512594972493e-05,
+      "loss": 3.9336,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9445,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9269,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9315,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660500156253788e-05,
+      "loss": 3.9388,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659661561502736e-05,
+      "loss": 3.9394,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588246046320574e-05,
+      "loss": 3.9179,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579860098810054e-05,
+      "loss": 3.9342,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657147415129953e-05,
+      "loss": 3.9222,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656308820378901e-05,
+      "loss": 3.9355,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655470225627849e-05,
+      "loss": 3.9327,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654631630876797e-05,
+      "loss": 3.9249,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653793036125745e-05,
+      "loss": 3.9387,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652954441374693e-05,
+      "loss": 3.9374,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521174845040136e-05,
+      "loss": 3.9265,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512788897529616e-05,
+      "loss": 3.933,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504402950019096e-05,
+      "loss": 3.9316,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649601700250858e-05,
+      "loss": 3.9299,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648763105499806e-05,
+      "loss": 3.9179,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647924510748754e-05,
+      "loss": 3.9327,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647085915997702e-05,
+      "loss": 3.9249,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646248959127023e-05,
+      "loss": 3.9311,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645412002256344e-05,
+      "loss": 3.9276,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644573407505292e-05,
+      "loss": 3.9299,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64373481275424e-05,
+      "loss": 3.9173,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642896218003188e-05,
+      "loss": 3.9233,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642057623252136e-05,
+      "loss": 3.9102,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641220666381457e-05,
+      "loss": 3.9367,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640382071630405e-05,
+      "loss": 3.9255,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395434768793536e-05,
+      "loss": 3.9249,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387048821283016e-05,
+      "loss": 3.9156,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378662873772496e-05,
+      "loss": 3.9314,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370276926261976e-05,
+      "loss": 3.9291,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361890978751456e-05,
+      "loss": 3.9171,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353505031240936e-05,
+      "loss": 3.9192,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345151841337874e-05,
+      "loss": 3.926,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336765893827354e-05,
+      "loss": 3.9219,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328379946316834e-05,
+      "loss": 3.9305,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319993998806314e-05,
+      "loss": 3.9125,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311624430099523e-05,
+      "loss": 3.9214,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303238482589003e-05,
+      "loss": 3.917,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629485253507848e-05,
+      "loss": 3.9114,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628646658756797e-05,
+      "loss": 3.9352,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627808064005745e-05,
+      "loss": 3.9133,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626971107135066e-05,
+      "loss": 3.9207,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626132512384014e-05,
+      "loss": 3.9189,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625293917632962e-05,
+      "loss": 3.9174,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9986412525177,
+      "eval_runtime": 302.0543,
+      "eval_samples_per_second": 1263.319,
+      "eval_steps_per_second": 39.48,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62445532288191e-05,
+      "loss": 3.9113,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623616728130858e-05,
+      "loss": 3.9116,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622778133379806e-05,
+      "loss": 3.9206,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621939538628753e-05,
+      "loss": 3.9163,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621102581758075e-05,
+      "loss": 3.9216,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620263987007023e-05,
+      "loss": 3.9032,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619427030136344e-05,
+      "loss": 3.9092,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185884353852924e-05,
+      "loss": 3.8974,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177498406342404e-05,
+      "loss": 3.9199,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169112458831884e-05,
+      "loss": 3.904,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616074289012509e-05,
+      "loss": 3.9119,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615235694261457e-05,
+      "loss": 3.9149,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614398737390778e-05,
+      "loss": 3.9037,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613560142639726e-05,
+      "loss": 3.8959,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612721547888674e-05,
+      "loss": 3.893,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611882953137622e-05,
+      "loss": 3.8932,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61104435838657e-05,
+      "loss": 3.9012,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610205763635518e-05,
+      "loss": 3.8978,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609367168884466e-05,
+      "loss": 3.8959,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608528574133414e-05,
+      "loss": 3.9198,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607691617262736e-05,
+      "loss": 3.9016,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606853022511684e-05,
+      "loss": 3.8988,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606014427760631e-05,
+      "loss": 3.9048,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605175833009579e-05,
+      "loss": 3.9034,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604337238258527e-05,
+      "loss": 3.8952,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603500281387848e-05,
+      "loss": 3.8898,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602661686636796e-05,
+      "loss": 3.9044,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601823091885744e-05,
+      "loss": 3.8854,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600984497134692e-05,
+      "loss": 3.8878,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.60014590238364e-05,
+      "loss": 3.8933,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993089455129615e-05,
+      "loss": 3.8874,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984703507619095e-05,
+      "loss": 3.9007,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.8932,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.8845,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959545665087535e-05,
+      "loss": 3.8975,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951176096380744e-05,
+      "loss": 3.9056,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942790148870224e-05,
+      "loss": 3.8818,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.893,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926018253849184e-05,
+      "loss": 3.8787,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917632306338664e-05,
+      "loss": 3.8928,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590926273763187e-05,
+      "loss": 3.8783,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590087679012135e-05,
+      "loss": 3.9012,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589249084261083e-05,
+      "loss": 3.891,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588410489510031e-05,
+      "loss": 3.8868,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587573532639353e-05,
+      "loss": 3.8917,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586734937888301e-05,
+      "loss": 3.88,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585896343137249e-05,
+      "loss": 3.8932,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585057748386197e-05,
+      "loss": 3.8773,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584220791515518e-05,
+      "loss": 3.8752,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583382196764466e-05,
+      "loss": 3.8905,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582543602013414e-05,
+      "loss": 3.8856,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581705007262362e-05,
+      "loss": 3.8859,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580868050391683e-05,
+      "loss": 3.8744,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580029455640631e-05,
+      "loss": 3.8835,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579190860889579e-05,
+      "loss": 3.8674,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783522661385267e-05,
+      "loss": 3.8887,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577513671387475e-05,
+      "loss": 3.882,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576676714516796e-05,
+      "loss": 3.895,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575838119765744e-05,
+      "loss": 3.8829,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574999525014692e-05,
+      "loss": 3.8757,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57416093026364e-05,
+      "loss": 3.8713,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573323973392961e-05,
+      "loss": 3.8823,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572485378641909e-05,
+      "loss": 3.8708,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571646783890857e-05,
+      "loss": 3.8688,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570808189139805e-05,
+      "loss": 3.8791,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569969594388753e-05,
+      "loss": 3.8758,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569130999637701e-05,
+      "loss": 3.8591,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568294042767022e-05,
+      "loss": 3.8646,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567455448015971e-05,
+      "loss": 3.8623,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566616853264919e-05,
+      "loss": 3.8871,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565778258513867e-05,
+      "loss": 3.8772,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.8706,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641027068921356e-05,
+      "loss": 3.8766,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632641121410836e-05,
+      "loss": 3.87,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624255173900316e-05,
+      "loss": 3.8824,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561586922638979e-05,
+      "loss": 3.8625,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607499657683005e-05,
+      "loss": 3.8831,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599113710172485e-05,
+      "loss": 3.8671,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590727762661965e-05,
+      "loss": 3.8738,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582341815151445e-05,
+      "loss": 3.8719,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573972246444654e-05,
+      "loss": 3.8697,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556558629893414e-05,
+      "loss": 3.8779,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557200351423614e-05,
+      "loss": 3.8694,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548814403913094e-05,
+      "loss": 3.8588,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554044483520631e-05,
+      "loss": 3.8688,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553205888769579e-05,
+      "loss": 3.8629,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552367294018526e-05,
+      "loss": 3.8733,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551528699267474e-05,
+      "loss": 3.8681,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550690104516422e-05,
+      "loss": 3.8537,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549853147645744e-05,
+      "loss": 3.8678,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549014552894691e-05,
+      "loss": 3.8594,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548175958143639e-05,
+      "loss": 3.8687,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547337363392588e-05,
+      "loss": 3.8555,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546500406521909e-05,
+      "loss": 3.8651,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545661811770857e-05,
+      "loss": 3.875,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544823217019805e-05,
+      "loss": 3.8596,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543984622268753e-05,
+      "loss": 3.8504,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431476653980736e-05,
+      "loss": 3.8606,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423090706470216e-05,
+      "loss": 3.8573,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414704758959696e-05,
+      "loss": 3.8718,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406318811449176e-05,
+      "loss": 3.8623,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397949242742385e-05,
+      "loss": 3.854,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389563295231865e-05,
+      "loss": 3.8553,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381177347721345e-05,
+      "loss": 3.8657,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537279140021083e-05,
+      "loss": 3.8444,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536442183150404e-05,
+      "loss": 3.856,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535603588399352e-05,
+      "loss": 3.858,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347649936483e-05,
+      "loss": 3.8623,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533926398897248e-05,
+      "loss": 3.8344,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533089442026569e-05,
+      "loss": 3.8563,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532250847275517e-05,
+      "loss": 3.8508,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531412252524465e-05,
+      "loss": 3.8495,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530573657773413e-05,
+      "loss": 3.8555,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529735063022361e-05,
+      "loss": 3.8444,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528898106151682e-05,
+      "loss": 3.8642,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52805951140063e-05,
+      "loss": 3.8557,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272209166495786e-05,
+      "loss": 3.8489,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263823218985266e-05,
+      "loss": 3.8557,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255453650278475e-05,
+      "loss": 3.8537,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247067702767955e-05,
+      "loss": 3.8543,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238681755257435e-05,
+      "loss": 3.836,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230295807746915e-05,
+      "loss": 3.8591,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221926239040124e-05,
+      "loss": 3.8451,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213540291529604e-05,
+      "loss": 3.852,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5205154344019084e-05,
+      "loss": 3.8504,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196768396508564e-05,
+      "loss": 3.8577,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518839882780177e-05,
+      "loss": 3.8463,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518001288029125e-05,
+      "loss": 3.8443,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517162693278074e-05,
+      "loss": 3.8369,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516324098527022e-05,
+      "loss": 3.8591,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515487141656343e-05,
+      "loss": 3.8519,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514648546905291e-05,
+      "loss": 3.851,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513809952154239e-05,
+      "loss": 3.8413,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512971357403187e-05,
+      "loss": 3.8562,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512134400532508e-05,
+      "loss": 3.8515,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511295805781456e-05,
+      "loss": 3.8423,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510457211030404e-05,
+      "loss": 3.8449,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509618616279352e-05,
+      "loss": 3.851,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087800215283e-05,
+      "loss": 3.8502,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079430646576206e-05,
+      "loss": 3.856,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507104469906569e-05,
+      "loss": 3.8377,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506265875155517e-05,
+      "loss": 3.848,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505427280404465e-05,
+      "loss": 3.8435,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504590323533786e-05,
+      "loss": 3.8421,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503753366663107e-05,
+      "loss": 3.8588,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502914771912055e-05,
+      "loss": 3.8428,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502076177161003e-05,
+      "loss": 3.8474,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501237582409951e-05,
+      "loss": 3.8413,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500398987658899e-05,
+      "loss": 3.8489,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9596714973449707,
+      "eval_runtime": 361.5509,
+      "eval_samples_per_second": 1055.428,
+      "eval_steps_per_second": 32.983,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499560392907847e-05,
+      "loss": 3.8323,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498721798156795e-05,
+      "loss": 3.8409,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497883203405743e-05,
+      "loss": 3.8482,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497044608654691e-05,
+      "loss": 3.8435,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496206013903639e-05,
+      "loss": 3.8541,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495367419152587e-05,
+      "loss": 3.8308,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494528824401535e-05,
+      "loss": 3.8386,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493690229650483e-05,
+      "loss": 3.8285,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928532727798047e-05,
+      "loss": 3.8448,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492014678028752e-05,
+      "loss": 3.8394,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911760832777e-05,
+      "loss": 3.8386,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490337488526648e-05,
+      "loss": 3.8394,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895021695363425e-05,
+      "loss": 3.8354,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886635747852905e-05,
+      "loss": 3.8263,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878249800342385e-05,
+      "loss": 3.8289,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869863852831864e-05,
+      "loss": 3.8184,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861477905321344e-05,
+      "loss": 3.834,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853091957810824e-05,
+      "loss": 3.8266,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844706010300304e-05,
+      "loss": 3.8229,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836320062789784e-05,
+      "loss": 3.8553,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827934115279264e-05,
+      "loss": 3.8295,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819548167768744e-05,
+      "loss": 3.8331,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4811162220258224e-05,
+      "loss": 3.8352,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480279265155143e-05,
+      "loss": 3.8303,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479440670404091e-05,
+      "loss": 3.8278,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478602075653039e-05,
+      "loss": 3.8231,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477763480901987e-05,
+      "loss": 3.8349,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476924886150935e-05,
+      "loss": 3.8126,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476087929280256e-05,
+      "loss": 3.8236,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475249334529205e-05,
+      "loss": 3.8219,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474410739778153e-05,
+      "loss": 3.8221,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473572145027101e-05,
+      "loss": 3.833,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472735188156422e-05,
+      "loss": 3.8282,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47189659340537e-05,
+      "loss": 3.8234,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471057998654318e-05,
+      "loss": 3.8295,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470219403903266e-05,
+      "loss": 3.8331,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469382447032587e-05,
+      "loss": 3.8167,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468543852281535e-05,
+      "loss": 3.828,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467705257530483e-05,
+      "loss": 3.8126,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466866662779431e-05,
+      "loss": 3.8271,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660297059087516e-05,
+      "loss": 3.8089,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651911111577e-05,
+      "loss": 3.8375,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464352516406648e-05,
+      "loss": 3.8248,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463513921655596e-05,
+      "loss": 3.8202,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462676964784917e-05,
+      "loss": 3.8262,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461838370033865e-05,
+      "loss": 3.818,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460999775282813e-05,
+      "loss": 3.8257,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460161180531761e-05,
+      "loss": 3.8138,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459324223661082e-05,
+      "loss": 3.8078,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45848562891003e-05,
+      "loss": 3.8268,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457647034158978e-05,
+      "loss": 3.8193,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456808439407926e-05,
+      "loss": 3.8215,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455969844656874e-05,
+      "loss": 3.8124,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551328877861956e-05,
+      "loss": 3.8179,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542942930351436e-05,
+      "loss": 3.8063,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534556982840916e-05,
+      "loss": 3.8226,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526171035330396e-05,
+      "loss": 3.817,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517801466623605e-05,
+      "loss": 3.8318,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509415519113085e-05,
+      "loss": 3.8201,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501029571602565e-05,
+      "loss": 3.8089,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492643624092045e-05,
+      "loss": 3.8125,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484274055385254e-05,
+      "loss": 3.819,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475888107874734e-05,
+      "loss": 3.8064,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467502160364214e-05,
+      "loss": 3.8086,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459116212853694e-05,
+      "loss": 3.8143,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445074664414691e-05,
+      "loss": 3.8167,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444236069663639e-05,
+      "loss": 3.7958,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443397474912587e-05,
+      "loss": 3.8045,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442558880161535e-05,
+      "loss": 3.7999,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441721923290856e-05,
+      "loss": 3.8197,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440883328539804e-05,
+      "loss": 3.815,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440044733788752e-05,
+      "loss": 3.8094,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4392061390377e-05,
+      "loss": 3.8138,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438369182167021e-05,
+      "loss": 3.8071,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437530587415969e-05,
+      "loss": 3.8248,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436691992664917e-05,
+      "loss": 3.801,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435853397913865e-05,
+      "loss": 3.8183,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350164410431864e-05,
+      "loss": 3.8097,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341778462921344e-05,
+      "loss": 3.8098,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333392515410824e-05,
+      "loss": 3.8121,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43250065679003e-05,
+      "loss": 3.8098,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431663699919351e-05,
+      "loss": 3.8149,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430825105168299e-05,
+      "loss": 3.81,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429986510417247e-05,
+      "loss": 3.7946,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291479156661946e-05,
+      "loss": 3.8092,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428310958795516e-05,
+      "loss": 3.8026,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427472364044464e-05,
+      "loss": 3.8128,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426633769293412e-05,
+      "loss": 3.8066,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42579517454236e-05,
+      "loss": 3.7949,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424958217671682e-05,
+      "loss": 3.8078,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42411962292063e-05,
+      "loss": 3.8023,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423281028169577e-05,
+      "loss": 3.8102,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422442433418525e-05,
+      "loss": 3.7934,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216054765478466e-05,
+      "loss": 3.8068,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207668817967946e-05,
+      "loss": 3.8202,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419928287045742e-05,
+      "loss": 3.7988,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41908969229469e-05,
+      "loss": 3.7861,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182527354240115e-05,
+      "loss": 3.7999,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174141406729595e-05,
+      "loss": 3.8034,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416575545921907e-05,
+      "loss": 3.8091,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157369511708555e-05,
+      "loss": 3.8053,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414899994300177e-05,
+      "loss": 3.7976,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140613995491244e-05,
+      "loss": 3.798,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132228047980724e-05,
+      "loss": 3.8082,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123842100470204e-05,
+      "loss": 3.7858,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411547253176342e-05,
+      "loss": 3.7969,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410708658425289e-05,
+      "loss": 3.7997,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409870063674237e-05,
+      "loss": 3.8036,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409031468923185e-05,
+      "loss": 3.7783,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408194512052507e-05,
+      "loss": 3.8,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407355917301454e-05,
+      "loss": 3.7909,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406517322550402e-05,
+      "loss": 3.796,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405678727799351e-05,
+      "loss": 3.7952,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404841770928672e-05,
+      "loss": 3.7902,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40400317617762e-05,
+      "loss": 3.8013,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403164581426568e-05,
+      "loss": 3.8027,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402325986675516e-05,
+      "loss": 3.7923,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401489029804837e-05,
+      "loss": 3.8006,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400650435053785e-05,
+      "loss": 3.7943,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399811840302733e-05,
+      "loss": 3.8001,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398973245551681e-05,
+      "loss": 3.7821,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981362886810016e-05,
+      "loss": 3.8043,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972976939299496e-05,
+      "loss": 3.7832,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964590991788976e-05,
+      "loss": 3.7971,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956205044278456e-05,
+      "loss": 3.7888,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.8032,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.7933,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393106358055063e-05,
+      "loss": 3.7884,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392267763304011e-05,
+      "loss": 3.7786,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.8074,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39059221168228e-05,
+      "loss": 3.789,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389753616931228e-05,
+      "loss": 3.7958,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388915022180176e-05,
+      "loss": 3.7855,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388078065309497e-05,
+      "loss": 3.8001,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387239470558445e-05,
+      "loss": 3.8003,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386400875807393e-05,
+      "loss": 3.7854,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385562281056341e-05,
+      "loss": 3.7877,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.7952,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838867294346105e-05,
+      "loss": 3.7955,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830481346835585e-05,
+      "loss": 3.8023,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822095399325065e-05,
+      "loss": 3.7801,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.7921,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7864,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796953935597234e-05,
+      "loss": 3.7914,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378858436689044e-05,
+      "loss": 3.8013,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378019841937992e-05,
+      "loss": 3.7835,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37718124718694e-05,
+      "loss": 3.7989,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376342652435888e-05,
+      "loss": 3.7837,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375504057684836e-05,
+      "loss": 3.7995,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9343461990356445,
+      "eval_runtime": 302.0015,
+      "eval_samples_per_second": 1263.54,
+      "eval_steps_per_second": 39.487,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374665462933785e-05,
+      "loss": 3.7801,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373826868182733e-05,
+      "loss": 3.7904,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372988273431681e-05,
+      "loss": 3.7898,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372151316561002e-05,
+      "loss": 3.7894,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37131272180995e-05,
+      "loss": 3.7974,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370474127058898e-05,
+      "loss": 3.7795,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369635532307846e-05,
+      "loss": 3.7823,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368798575437167e-05,
+      "loss": 3.7754,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367959980686115e-05,
+      "loss": 3.7921,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367121385935063e-05,
+      "loss": 3.7887,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366282791184011e-05,
+      "loss": 3.7833,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365445834313332e-05,
+      "loss": 3.7882,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364608877442653e-05,
+      "loss": 3.779,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363770282691601e-05,
+      "loss": 3.775,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362931687940549e-05,
+      "loss": 3.7784,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362093093189497e-05,
+      "loss": 3.7655,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361254498438445e-05,
+      "loss": 3.7797,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360415903687393e-05,
+      "loss": 3.7751,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359577308936341e-05,
+      "loss": 3.772,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358738714185289e-05,
+      "loss": 3.8093,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35790175731461e-05,
+      "loss": 3.7757,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357063162563558e-05,
+      "loss": 3.783,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356224567812506e-05,
+      "loss": 3.7816,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355387610941827e-05,
+      "loss": 3.7807,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354549016190776e-05,
+      "loss": 3.7734,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353710421439724e-05,
+      "loss": 3.7758,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352871826688672e-05,
+      "loss": 3.7813,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520348698179926e-05,
+      "loss": 3.7635,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511962750669406e-05,
+      "loss": 3.7721,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503576803158886e-05,
+      "loss": 3.7686,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495190855648366e-05,
+      "loss": 3.7739,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486821286941575e-05,
+      "loss": 3.7772,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478435339431055e-05,
+      "loss": 3.7771,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470049391920535e-05,
+      "loss": 3.7728,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461663444410015e-05,
+      "loss": 3.7786,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453293875703224e-05,
+      "loss": 3.7803,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344490792819271e-05,
+      "loss": 3.7663,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343652198068219e-05,
+      "loss": 3.7795,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342813603317167e-05,
+      "loss": 3.7604,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341976646446488e-05,
+      "loss": 3.7803,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341139689575809e-05,
+      "loss": 3.7548,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340301094824757e-05,
+      "loss": 3.7912,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339462500073705e-05,
+      "loss": 3.7738,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338623905322653e-05,
+      "loss": 3.7706,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337785310571601e-05,
+      "loss": 3.7749,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336946715820549e-05,
+      "loss": 3.7721,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336108121069497e-05,
+      "loss": 3.7725,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335269526318445e-05,
+      "loss": 3.7604,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344325694477664e-05,
+      "loss": 3.7521,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335939746967144e-05,
+      "loss": 3.7827,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327553799456624e-05,
+      "loss": 3.7705,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33191678519461e-05,
+      "loss": 3.7764,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331079828323931e-05,
+      "loss": 3.7615,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330241233572879e-05,
+      "loss": 3.7671,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294026388218266e-05,
+      "loss": 3.7568,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285640440707746e-05,
+      "loss": 3.7721,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327727087200096e-05,
+      "loss": 3.7678,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326888492449044e-05,
+      "loss": 3.7822,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260498976979915e-05,
+      "loss": 3.7743,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32521130294694e-05,
+      "loss": 3.7584,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324374346076261e-05,
+      "loss": 3.7623,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32353575132521e-05,
+      "loss": 3.771,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322697156574157e-05,
+      "loss": 3.7573,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321858561823105e-05,
+      "loss": 3.7586,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321021604952427e-05,
+      "loss": 3.7678,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320183010201374e-05,
+      "loss": 3.7653,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319344415450322e-05,
+      "loss": 3.7496,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31850582069927e-05,
+      "loss": 3.7564,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176688638285916e-05,
+      "loss": 3.7508,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316830269077539e-05,
+      "loss": 3.7711,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315991674326487e-05,
+      "loss": 3.7657,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151530795754356e-05,
+      "loss": 3.7617,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3143161227047565e-05,
+      "loss": 3.7686,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134775279537045e-05,
+      "loss": 3.7571,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126389332026525e-05,
+      "loss": 3.7775,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311801976331974e-05,
+      "loss": 3.754,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109633815809214e-05,
+      "loss": 3.7682,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3101247868298694e-05,
+      "loss": 3.7625,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092861920788174e-05,
+      "loss": 3.7599,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7665,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307610640457086e-05,
+      "loss": 3.7623,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306772045706034e-05,
+      "loss": 3.7654,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305933450954982e-05,
+      "loss": 3.7628,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7487,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304257899333252e-05,
+      "loss": 3.7633,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034193045822e-05,
+      "loss": 3.7556,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302580709831148e-05,
+      "loss": 3.7663,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301743752960469e-05,
+      "loss": 3.7556,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300905158209417e-05,
+      "loss": 3.7495,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300066563458365e-05,
+      "loss": 3.7594,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299227968707313e-05,
+      "loss": 3.7601,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983910118366337e-05,
+      "loss": 3.7595,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975524170855817e-05,
+      "loss": 3.7482,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2967138223345296e-05,
+      "loss": 3.7638,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958752275834776e-05,
+      "loss": 3.77,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295038270712799e-05,
+      "loss": 3.7565,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294199675961747e-05,
+      "loss": 3.7383,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293361081210695e-05,
+      "loss": 3.7555,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292522486459643e-05,
+      "loss": 3.7563,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7595,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.762,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29000834008686e-05,
+      "loss": 3.7533,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289169745335808e-05,
+      "loss": 3.7508,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7605,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.7419,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286655598963025e-05,
+      "loss": 3.7522,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285817004211973e-05,
+      "loss": 3.7497,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.7597,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841414525902426e-05,
+      "loss": 3.7361,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833028578391906e-05,
+      "loss": 3.7484,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824642630881386e-05,
+      "loss": 3.7488,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.7487,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.751,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7421,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791115219643035e-05,
+      "loss": 3.7578,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782745650936244e-05,
+      "loss": 3.7578,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774359703425724e-05,
+      "loss": 3.7443,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765973755915204e-05,
+      "loss": 3.7582,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757587808404684e-05,
+      "loss": 3.7456,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27492182396979e-05,
+      "loss": 3.756,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7407,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7552,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272406039716634e-05,
+      "loss": 3.7374,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271569082845955e-05,
+      "loss": 3.7514,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270730488094903e-05,
+      "loss": 3.7468,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269891893343851e-05,
+      "loss": 3.7525,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269053298592799e-05,
+      "loss": 3.7474,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26821634172212e-05,
+      "loss": 3.7483,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267377746971068e-05,
+      "loss": 3.7312,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266539152220016e-05,
+      "loss": 3.7649,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265700557468964e-05,
+      "loss": 3.7432,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264863600598285e-05,
+      "loss": 3.7526,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264025005847233e-05,
+      "loss": 3.7392,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263186411096181e-05,
+      "loss": 3.7559,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262347816345129e-05,
+      "loss": 3.7526,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26151085947445e-05,
+      "loss": 3.7428,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260672264723398e-05,
+      "loss": 3.7466,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259833669972346e-05,
+      "loss": 3.75,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258995075221294e-05,
+      "loss": 3.7507,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258158118350615e-05,
+      "loss": 3.7653,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257319523599563e-05,
+      "loss": 3.7354,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256480928848511e-05,
+      "loss": 3.7535,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255642334097459e-05,
+      "loss": 3.7389,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254805377226781e-05,
+      "loss": 3.7514,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539684203561016e-05,
+      "loss": 3.7543,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2531298256050496e-05,
+      "loss": 3.7398,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522912308539976e-05,
+      "loss": 3.7568,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514526361029456e-05,
+      "loss": 3.7449,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2506140413518936e-05,
+      "loss": 3.7523,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.918186902999878,
+      "eval_runtime": 302.0856,
+      "eval_samples_per_second": 1263.188,
+      "eval_steps_per_second": 39.476,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497754466008416e-05,
+      "loss": 3.7358,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489368518497896e-05,
+      "loss": 3.7421,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480982570987376e-05,
+      "loss": 3.7495,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472596623476856e-05,
+      "loss": 3.7459,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464210675966336e-05,
+      "loss": 3.7514,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455824728455816e-05,
+      "loss": 3.7415,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447455159749025e-05,
+      "loss": 3.7342,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243906921223851e-05,
+      "loss": 3.7372,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243068326472799e-05,
+      "loss": 3.7469,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242229731721747e-05,
+      "loss": 3.748,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413911369706945e-05,
+      "loss": 3.7405,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405525422196424e-05,
+      "loss": 3.7472,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239715585348964e-05,
+      "loss": 3.7298,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388769905979114e-05,
+      "loss": 3.7372,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380383958468593e-05,
+      "loss": 3.7371,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371998010958073e-05,
+      "loss": 3.7233,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363612063447553e-05,
+      "loss": 3.7379,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235522611593703e-05,
+      "loss": 3.7284,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234684016842651e-05,
+      "loss": 3.7294,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233845422091599e-05,
+      "loss": 3.7686,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233008465220921e-05,
+      "loss": 3.7353,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232169870469869e-05,
+      "loss": 3.7383,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231331275718817e-05,
+      "loss": 3.7382,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230494318848138e-05,
+      "loss": 3.7446,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229655724097086e-05,
+      "loss": 3.7228,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228817129346034e-05,
+      "loss": 3.7361,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227978534594982e-05,
+      "loss": 3.7373,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227141577724303e-05,
+      "loss": 3.7266,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226302982973251e-05,
+      "loss": 3.7308,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254660261025716e-05,
+      "loss": 3.7249,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246274313515196e-05,
+      "loss": 3.7357,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223788836600468e-05,
+      "loss": 3.7374,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222950241849416e-05,
+      "loss": 3.7345,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222111647098364e-05,
+      "loss": 3.7304,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221273052347312e-05,
+      "loss": 3.7349,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22043445759626e-05,
+      "loss": 3.7394,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219597500725581e-05,
+      "loss": 3.725,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218758905974529e-05,
+      "loss": 3.7421,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217920311223477e-05,
+      "loss": 3.7185,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217081716472425e-05,
+      "loss": 3.7367,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216244759601746e-05,
+      "loss": 3.7148,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215406164850694e-05,
+      "loss": 3.7489,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214567570099642e-05,
+      "loss": 3.731,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21372897534859e-05,
+      "loss": 3.7334,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212890380597538e-05,
+      "loss": 3.7328,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212051785846487e-05,
+      "loss": 3.7311,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211213191095435e-05,
+      "loss": 3.7346,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210374596344383e-05,
+      "loss": 3.7215,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095376394737036e-05,
+      "loss": 3.7086,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086990447226516e-05,
+      "loss": 3.7428,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078604499715996e-05,
+      "loss": 3.7263,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070218552205476e-05,
+      "loss": 3.7387,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061848983498685e-05,
+      "loss": 3.7216,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053463035988165e-05,
+      "loss": 3.7284,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045077088477645e-05,
+      "loss": 3.7147,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036691140967125e-05,
+      "loss": 3.7347,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028321572260334e-05,
+      "loss": 3.7303,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201993562474982e-05,
+      "loss": 3.7383,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20115496772393e-05,
+      "loss": 3.7347,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200316372972878e-05,
+      "loss": 3.717,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199479416102199e-05,
+      "loss": 3.7221,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198640821351147e-05,
+      "loss": 3.7383,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197802226600095e-05,
+      "loss": 3.7151,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196963631849042e-05,
+      "loss": 3.7196,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196126674978364e-05,
+      "loss": 3.7263,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195288080227312e-05,
+      "loss": 3.7318,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19444948547626e-05,
+      "loss": 3.7089,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193610890725207e-05,
+      "loss": 3.7165,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192772295974156e-05,
+      "loss": 3.715,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191933701223104e-05,
+      "loss": 3.7228,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191095106472052e-05,
+      "loss": 3.7295,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190258149601373e-05,
+      "loss": 3.7237,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189419554850321e-05,
+      "loss": 3.7289,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188580960099269e-05,
+      "loss": 3.7157,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187742365348217e-05,
+      "loss": 3.741,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186905408477538e-05,
+      "loss": 3.716,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186066813726486e-05,
+      "loss": 3.729,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185228218975434e-05,
+      "loss": 3.7231,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843896242243817e-05,
+      "loss": 3.7202,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183554305234076e-05,
+      "loss": 3.7275,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182715710483024e-05,
+      "loss": 3.7255,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181877115731972e-05,
+      "loss": 3.7217,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18103852098092e-05,
+      "loss": 3.7277,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180199926229868e-05,
+      "loss": 3.7119,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179361331478816e-05,
+      "loss": 3.7209,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178522736727764e-05,
+      "loss": 3.7215,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177684141976712e-05,
+      "loss": 3.7259,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176847185106033e-05,
+      "loss": 3.7175,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176008590354981e-05,
+      "loss": 3.7113,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175169995603929e-05,
+      "loss": 3.717,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174331400852877e-05,
+      "loss": 3.7256,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173494443982198e-05,
+      "loss": 3.7216,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726558492311466e-05,
+      "loss": 3.709,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718172544800946e-05,
+      "loss": 3.7236,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709786597290426e-05,
+      "loss": 3.7296,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701417028583635e-05,
+      "loss": 3.7206,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693031081073115e-05,
+      "loss": 3.7011,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684645133562595e-05,
+      "loss": 3.7172,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1676259186052075e-05,
+      "loss": 3.7125,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667889617345284e-05,
+      "loss": 3.7254,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659503669834764e-05,
+      "loss": 3.7267,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651117722324244e-05,
+      "loss": 3.7129,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642731774813724e-05,
+      "loss": 3.714,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163436220610693e-05,
+      "loss": 3.723,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162597625859642e-05,
+      "loss": 3.7039,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16175903110859e-05,
+      "loss": 3.7173,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160920436357538e-05,
+      "loss": 3.709,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160083479486859e-05,
+      "loss": 3.7275,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159244884735807e-05,
+      "loss": 3.6971,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158406289984755e-05,
+      "loss": 3.7092,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157567695233703e-05,
+      "loss": 3.7102,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156732376243397e-05,
+      "loss": 3.7153,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155893781492345e-05,
+      "loss": 3.7104,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155055186741293e-05,
+      "loss": 3.7071,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154216591990241e-05,
+      "loss": 3.717,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153377997239189e-05,
+      "loss": 3.7213,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152539402488137e-05,
+      "loss": 3.7079,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151700807737085e-05,
+      "loss": 3.724,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150862212986033e-05,
+      "loss": 3.7086,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150025256115354e-05,
+      "loss": 3.7147,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149186661364302e-05,
+      "loss": 3.7057,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14834806661325e-05,
+      "loss": 3.7142,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147509471862198e-05,
+      "loss": 3.7025,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146672514991519e-05,
+      "loss": 3.713,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145833920240467e-05,
+      "loss": 3.7153,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144995325489415e-05,
+      "loss": 3.7116,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144156730738363e-05,
+      "loss": 3.7098,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143319773867684e-05,
+      "loss": 3.7135,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142481179116632e-05,
+      "loss": 3.6966,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141642584365581e-05,
+      "loss": 3.7246,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1408056274949016e-05,
+      "loss": 3.71,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399670327438496e-05,
+      "loss": 3.7157,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391284379927976e-05,
+      "loss": 3.7047,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382898432417456e-05,
+      "loss": 3.7238,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374528863710665e-05,
+      "loss": 3.7121,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366142916200145e-05,
+      "loss": 3.7054,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357756968689625e-05,
+      "loss": 3.7086,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349371021179105e-05,
+      "loss": 3.7121,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1341001452472314e-05,
+      "loss": 3.7142,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133263188376553e-05,
+      "loss": 3.7271,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132424593625501e-05,
+      "loss": 3.6987,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131585998874449e-05,
+      "loss": 3.7154,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130747404123397e-05,
+      "loss": 3.7043,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129908809372345e-05,
+      "loss": 3.7162,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129071852501666e-05,
+      "loss": 3.7141,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128233257750614e-05,
+      "loss": 3.7059,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127394662999562e-05,
+      "loss": 3.7176,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12655606824851e-05,
+      "loss": 3.7085,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125717473497458e-05,
+      "loss": 3.7141,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9086039066314697,
+      "eval_runtime": 303.79,
+      "eval_samples_per_second": 1256.101,
+      "eval_steps_per_second": 39.254,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124878878746406e-05,
+      "loss": 3.7077,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124040283995354e-05,
+      "loss": 3.7029,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123201689244302e-05,
+      "loss": 3.7102,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122364732373623e-05,
+      "loss": 3.7118,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1215261376225714e-05,
+      "loss": 3.7174,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206875428715194e-05,
+      "loss": 3.7028,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1198505860008403e-05,
+      "loss": 3.6993,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119011991249788e-05,
+      "loss": 3.6972,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118173396498736e-05,
+      "loss": 3.7121,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117334801747684e-05,
+      "loss": 3.7138,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116496206996632e-05,
+      "loss": 3.6986,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156576122455796e-05,
+      "loss": 3.7106,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114820655374901e-05,
+      "loss": 3.7003,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113982060623849e-05,
+      "loss": 3.7012,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113143465872797e-05,
+      "loss": 3.6972,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112304871121745e-05,
+      "loss": 3.6917,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111466276370693e-05,
+      "loss": 3.6996,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110627681619641e-05,
+      "loss": 3.6962,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109789086868589e-05,
+      "loss": 3.6925,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108950492117537e-05,
+      "loss": 3.7275,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108113535246858e-05,
+      "loss": 3.7046,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107274940495806e-05,
+      "loss": 3.7021,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106436345744754e-05,
+      "loss": 3.7017,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105597750993702e-05,
+      "loss": 3.7087,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104760794123023e-05,
+      "loss": 3.6899,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103922199371971e-05,
+      "loss": 3.7025,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103085242501292e-05,
+      "loss": 3.7033,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1022466477502406e-05,
+      "loss": 3.6939,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1014080529991886e-05,
+      "loss": 3.693,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005694582481366e-05,
+      "loss": 3.6901,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997308634970846e-05,
+      "loss": 3.6994,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988922687460326e-05,
+      "loss": 3.7034,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980536739949806e-05,
+      "loss": 3.7011,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972150792439286e-05,
+      "loss": 3.6957,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963781223732495e-05,
+      "loss": 3.7004,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955395276221975e-05,
+      "loss": 3.702,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0947009328711455e-05,
+      "loss": 3.6963,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938623381200935e-05,
+      "loss": 3.7039,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093027019129787e-05,
+      "loss": 3.6883,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092188424378736e-05,
+      "loss": 3.7002,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091349829627684e-05,
+      "loss": 3.6852,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090511234876632e-05,
+      "loss": 3.7139,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089674278005953e-05,
+      "loss": 3.6983,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088835683254901e-05,
+      "loss": 3.6975,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087997088503849e-05,
+      "loss": 3.7015,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087158493752797e-05,
+      "loss": 3.6954,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086319899001745e-05,
+      "loss": 3.7002,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085481304250693e-05,
+      "loss": 3.6889,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084642709499641e-05,
+      "loss": 3.6705,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083804114748589e-05,
+      "loss": 3.7092,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08296715787791e-05,
+      "loss": 3.693,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082130201007231e-05,
+      "loss": 3.7058,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081291606256179e-05,
+      "loss": 3.6889,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080453011505127e-05,
+      "loss": 3.6924,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079614416754075e-05,
+      "loss": 3.6804,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078775822003023e-05,
+      "loss": 3.6995,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077937227251971e-05,
+      "loss": 3.6967,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077098632500919e-05,
+      "loss": 3.7008,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076260037749867e-05,
+      "loss": 3.7054,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075423080879188e-05,
+      "loss": 3.6856,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074584486128136e-05,
+      "loss": 3.6882,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073745891377084e-05,
+      "loss": 3.704,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072907296626032e-05,
+      "loss": 3.6838,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072070339755353e-05,
+      "loss": 3.6821,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071233382884675e-05,
+      "loss": 3.6941,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070394788133623e-05,
+      "loss": 3.695,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069556193382571e-05,
+      "loss": 3.6801,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068717598631519e-05,
+      "loss": 3.6839,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678790038804667e-05,
+      "loss": 3.679,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0670404091294147e-05,
+      "loss": 3.6903,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0662018143783627e-05,
+      "loss": 3.6959,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653648575076836e-05,
+      "loss": 3.693,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645262627566316e-05,
+      "loss": 3.6899,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636876680055796e-05,
+      "loss": 3.6875,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628490732545275e-05,
+      "loss": 3.707,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620104785034755e-05,
+      "loss": 3.6806,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061171883752423e-05,
+      "loss": 3.6994,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0603332890013715e-05,
+      "loss": 3.6885,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594946942503195e-05,
+      "loss": 3.6836,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586577373796404e-05,
+      "loss": 3.6978,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578191426285884e-05,
+      "loss": 3.694,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0569805478775364e-05,
+      "loss": 3.6897,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0561419531264844e-05,
+      "loss": 3.6946,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055304996255805e-05,
+      "loss": 3.6798,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054466401504753e-05,
+      "loss": 3.6861,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053627806753701e-05,
+      "loss": 3.688,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052790849883023e-05,
+      "loss": 3.6944,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05195225513197e-05,
+      "loss": 3.6826,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051113660380918e-05,
+      "loss": 3.6821,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050275065629867e-05,
+      "loss": 3.6803,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049438108759188e-05,
+      "loss": 3.6933,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048599514008136e-05,
+      "loss": 3.6906,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047760919257084e-05,
+      "loss": 3.679,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046922324506032e-05,
+      "loss": 3.6927,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046085367635353e-05,
+      "loss": 3.6968,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045246772884301e-05,
+      "loss": 3.6865,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044408178133249e-05,
+      "loss": 3.6696,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043569583382197e-05,
+      "loss": 3.6852,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427326265115176e-05,
+      "loss": 3.6796,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418940317604656e-05,
+      "loss": 3.6944,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0410554370094136e-05,
+      "loss": 3.6933,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040216842258362e-05,
+      "loss": 3.6824,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039379885387683e-05,
+      "loss": 3.6839,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038541290636631e-05,
+      "loss": 3.6859,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037702695885579e-05,
+      "loss": 3.6775,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036864101134527e-05,
+      "loss": 3.6856,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036027144263848e-05,
+      "loss": 3.6761,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035188549512796e-05,
+      "loss": 3.6988,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0343515926421177e-05,
+      "loss": 3.665,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033512997891065e-05,
+      "loss": 3.6763,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032674403140013e-05,
+      "loss": 3.6766,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031835808388961e-05,
+      "loss": 3.6879,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030997213637909e-05,
+      "loss": 3.6751,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301586188868576e-05,
+      "loss": 3.676,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293200241358056e-05,
+      "loss": 3.6858,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284830672651265e-05,
+      "loss": 3.6886,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276444725140745e-05,
+      "loss": 3.673,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0268058777630225e-05,
+      "loss": 3.688,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259672830119705e-05,
+      "loss": 3.6817,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0251303261412914e-05,
+      "loss": 3.683,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242917313902394e-05,
+      "loss": 3.6767,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0234531366391874e-05,
+      "loss": 3.6785,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0226145418881354e-05,
+      "loss": 3.6739,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217759471370834e-05,
+      "loss": 3.6835,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020938990266404e-05,
+      "loss": 3.6855,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020100395515353e-05,
+      "loss": 3.6812,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019261800764301e-05,
+      "loss": 3.6788,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018423206013249e-05,
+      "loss": 3.6801,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017584611262197e-05,
+      "loss": 3.6654,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016746016511145e-05,
+      "loss": 3.6957,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015907421760093e-05,
+      "loss": 3.6773,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015070464889414e-05,
+      "loss": 3.6843,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014231870138362e-05,
+      "loss": 3.6722,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01339327538731e-05,
+      "loss": 3.6905,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012554680636258e-05,
+      "loss": 3.6825,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011716085885206e-05,
+      "loss": 3.6748,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010877491134154e-05,
+      "loss": 3.6795,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010038896383101e-05,
+      "loss": 3.6775,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009201939512423e-05,
+      "loss": 3.6827,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008363344761371e-05,
+      "loss": 3.7005,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007524750010319e-05,
+      "loss": 3.6634,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006686155259267e-05,
+      "loss": 3.6883,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0058491983885883e-05,
+      "loss": 3.6728,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0050106036375363e-05,
+      "loss": 3.6824,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004173646766857e-05,
+      "loss": 3.6784,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003335052015805e-05,
+      "loss": 3.6804,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002496457264753e-05,
+      "loss": 3.6842,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016578625137006e-05,
+      "loss": 3.6787,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0008192677626486e-05,
+      "loss": 3.6827,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.901306629180908,
+      "eval_runtime": 303.9595,
+      "eval_samples_per_second": 1255.401,
+      "eval_steps_per_second": 39.232,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999806730115966e-05,
+      "loss": 3.676,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991420782605445e-05,
+      "loss": 3.6724,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998303483509493e-05,
+      "loss": 3.6811,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997464888758441e-05,
+      "loss": 3.6797,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996626294007389e-05,
+      "loss": 3.6892,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995787699256337e-05,
+      "loss": 3.6709,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994950742385658e-05,
+      "loss": 3.6703,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994112147634606e-05,
+      "loss": 3.6732,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993273552883554e-05,
+      "loss": 3.6774,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992434958132502e-05,
+      "loss": 3.6818,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99159636338145e-05,
+      "loss": 3.6705,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990757768630398e-05,
+      "loss": 3.6772,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989920811759719e-05,
+      "loss": 3.6756,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989082217008667e-05,
+      "loss": 3.6664,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988243622257615e-05,
+      "loss": 3.6725,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987405027506563e-05,
+      "loss": 3.6601,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986566432755512e-05,
+      "loss": 3.6687,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857278380044597e-05,
+      "loss": 3.6648,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848892432534077e-05,
+      "loss": 3.6641,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984050648502355e-05,
+      "loss": 3.695,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9832136916316766e-05,
+      "loss": 3.6782,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823750968806246e-05,
+      "loss": 3.6727,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815365021295726e-05,
+      "loss": 3.6727,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98069790737852e-05,
+      "loss": 3.6777,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9798609505078415e-05,
+      "loss": 3.657,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790223557567895e-05,
+      "loss": 3.6739,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781853988861104e-05,
+      "loss": 3.6735,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773468041350584e-05,
+      "loss": 3.661,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976508209384007e-05,
+      "loss": 3.6614,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975669614632955e-05,
+      "loss": 3.6627,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748310198819023e-05,
+      "loss": 3.6686,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739924251308503e-05,
+      "loss": 3.6746,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973153830379798e-05,
+      "loss": 3.6702,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972315235628746e-05,
+      "loss": 3.669,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971478278758067e-05,
+      "loss": 3.672,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970639684007015e-05,
+      "loss": 3.6721,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969801089255963e-05,
+      "loss": 3.6696,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968962494504911e-05,
+      "loss": 3.6716,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968125537634232e-05,
+      "loss": 3.6565,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967288580763554e-05,
+      "loss": 3.6689,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9664499860125024e-05,
+      "loss": 3.6577,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96561139126145e-05,
+      "loss": 3.6827,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964772796510398e-05,
+      "loss": 3.6725,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963934201759346e-05,
+      "loss": 3.667,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963095607008294e-05,
+      "loss": 3.6751,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962257012257242e-05,
+      "loss": 3.6657,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96141841750619e-05,
+      "loss": 3.6736,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960583098515884e-05,
+      "loss": 3.6609,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9597445037648315e-05,
+      "loss": 3.6364,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9589059090137795e-05,
+      "loss": 3.6793,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9580673142627275e-05,
+      "loss": 3.6643,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957228719511676e-05,
+      "loss": 3.6801,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956390124760624e-05,
+      "loss": 3.6623,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955551530009572e-05,
+      "loss": 3.6607,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95471293525852e-05,
+      "loss": 3.652,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953877616268215e-05,
+      "loss": 3.6662,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953039021517162e-05,
+      "loss": 3.6735,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95220042676611e-05,
+      "loss": 3.6695,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951361832015058e-05,
+      "loss": 3.6777,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950523237264006e-05,
+      "loss": 3.6585,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949684642512954e-05,
+      "loss": 3.6552,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948846047761902e-05,
+      "loss": 3.6753,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94800745301085e-05,
+      "loss": 3.6544,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947170496140171e-05,
+      "loss": 3.6558,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9463335392694925e-05,
+      "loss": 3.6625,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454949445184404e-05,
+      "loss": 3.67,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446563497673884e-05,
+      "loss": 3.6503,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438177550163364e-05,
+      "loss": 3.6575,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429807981456573e-05,
+      "loss": 3.6515,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9421422033946053e-05,
+      "loss": 3.6575,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9413036086435533e-05,
+      "loss": 3.6685,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940465013892501e-05,
+      "loss": 3.6597,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939626419141449e-05,
+      "loss": 3.6633,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938787824390397e-05,
+      "loss": 3.6569,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937949229639345e-05,
+      "loss": 3.6812,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937112272768666e-05,
+      "loss": 3.6547,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936273678017615e-05,
+      "loss": 3.6691,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935435083266563e-05,
+      "loss": 3.6594,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934596488515511e-05,
+      "loss": 3.6549,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933759531644832e-05,
+      "loss": 3.6683,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93292093689378e-05,
+      "loss": 3.667,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932082342142728e-05,
+      "loss": 3.6632,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931243747391676e-05,
+      "loss": 3.6616,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930405152640624e-05,
+      "loss": 3.6559,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929566557889572e-05,
+      "loss": 3.6546,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92872796313852e-05,
+      "loss": 3.6622,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927889368387468e-05,
+      "loss": 3.6663,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270540493971616e-05,
+      "loss": 3.6551,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92621545464611e-05,
+      "loss": 3.6543,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925376859895058e-05,
+      "loss": 3.6525,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924538265144006e-05,
+      "loss": 3.6679,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923699670392954e-05,
+      "loss": 3.661,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922861075641902e-05,
+      "loss": 3.651,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92202248089085e-05,
+      "loss": 3.6641,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211838861397976e-05,
+      "loss": 3.6677,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920346929269119e-05,
+      "loss": 3.659,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919508334518067e-05,
+      "loss": 3.644,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918669739767015e-05,
+      "loss": 3.6553,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9178311450159625e-05,
+      "loss": 3.6513,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916994188145284e-05,
+      "loss": 3.6676,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916155593394232e-05,
+      "loss": 3.6633,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91531699864318e-05,
+      "loss": 3.6615,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914478403892128e-05,
+      "loss": 3.6536,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913639809141076e-05,
+      "loss": 3.6566,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9128028522703976e-05,
+      "loss": 3.651,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911964257519345e-05,
+      "loss": 3.6561,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911125662768293e-05,
+      "loss": 3.6481,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910287068017241e-05,
+      "loss": 3.6684,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094501111465625e-05,
+      "loss": 3.6411,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90861151639551e-05,
+      "loss": 3.6447,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907772921644458e-05,
+      "loss": 3.6518,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906934326893406e-05,
+      "loss": 3.6531,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906095732142354e-05,
+      "loss": 3.6519,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9052571373913025e-05,
+      "loss": 3.6485,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9044185426402505e-05,
+      "loss": 3.657,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035815857695714e-05,
+      "loss": 3.665,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027429910185194e-05,
+      "loss": 3.644,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9019043962674674e-05,
+      "loss": 3.6575,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9010658015164154e-05,
+      "loss": 3.6553,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9002272067653634e-05,
+      "loss": 3.6579,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993886120143114e-05,
+      "loss": 3.6515,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8985500172632594e-05,
+      "loss": 3.6499,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8977114225122074e-05,
+      "loss": 3.6479,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896876103521901e-05,
+      "loss": 3.654,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896037508770849e-05,
+      "loss": 3.6582,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895198914019798e-05,
+      "loss": 3.652,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894360319268746e-05,
+      "loss": 3.6539,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893521724517694e-05,
+      "loss": 3.6505,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892684767647015e-05,
+      "loss": 3.6389,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891846172895963e-05,
+      "loss": 3.6674,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891007578144911e-05,
+      "loss": 3.6473,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890168983393859e-05,
+      "loss": 3.6604,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889330388642807e-05,
+      "loss": 3.6434,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888491793891755e-05,
+      "loss": 3.6629,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887653199140703e-05,
+      "loss": 3.6561,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886814604389651e-05,
+      "loss": 3.6454,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859776475189716e-05,
+      "loss": 3.6505,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885140690648293e-05,
+      "loss": 3.6506,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884302095897241e-05,
+      "loss": 3.6566,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883463501146189e-05,
+      "loss": 3.67,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882624906395137e-05,
+      "loss": 3.6388,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881786311644085e-05,
+      "loss": 3.6623,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880949354773406e-05,
+      "loss": 3.6434,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880110760022354e-05,
+      "loss": 3.6568,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879272165271302e-05,
+      "loss": 3.6543,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87843357052025e-05,
+      "loss": 3.6508,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877594975769198e-05,
+      "loss": 3.6576,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876756381018146e-05,
+      "loss": 3.656,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875919424147467e-05,
+      "loss": 3.6528,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8980462551116943,
+      "eval_runtime": 303.9561,
+      "eval_samples_per_second": 1255.415,
+      "eval_steps_per_second": 39.233,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875080829396415e-05,
+      "loss": 3.6481,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874242234645363e-05,
+      "loss": 3.6477,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873403639894311e-05,
+      "loss": 3.6536,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872565045143259e-05,
+      "loss": 3.651,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871726450392207e-05,
+      "loss": 3.6611,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870887855641155e-05,
+      "loss": 3.645,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870050898770476e-05,
+      "loss": 3.6442,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869212304019424e-05,
+      "loss": 3.6458,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868373709268372e-05,
+      "loss": 3.6501,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86753511451732e-05,
+      "loss": 3.6539,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866696519766268e-05,
+      "loss": 3.6467,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865859562895589e-05,
+      "loss": 3.6492,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8650226060249104e-05,
+      "loss": 3.6489,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641840112738584e-05,
+      "loss": 3.6398,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8633454165228064e-05,
+      "loss": 3.6436,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8625068217717544e-05,
+      "loss": 3.6357,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8616682270207024e-05,
+      "loss": 3.6386,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8608296322696503e-05,
+      "loss": 3.6415,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599910375185983e-05,
+      "loss": 3.6386,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8591524427675463e-05,
+      "loss": 3.6623,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858313848016494e-05,
+      "loss": 3.6539,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857475253265442e-05,
+      "loss": 3.6482,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85663665851439e-05,
+      "loss": 3.6457,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855798063763338e-05,
+      "loss": 3.6513,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854961106892659e-05,
+      "loss": 3.6305,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854122512141607e-05,
+      "loss": 3.6466,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853285555270929e-05,
+      "loss": 3.6492,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852446960519877e-05,
+      "loss": 3.6301,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851608365768825e-05,
+      "loss": 3.6415,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850769771017773e-05,
+      "loss": 3.634,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849931176266721e-05,
+      "loss": 3.6447,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849092581515669e-05,
+      "loss": 3.6432,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848253986764617e-05,
+      "loss": 3.6458,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847415392013565e-05,
+      "loss": 3.6513,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846578435142886e-05,
+      "loss": 3.6401,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845739840391834e-05,
+      "loss": 3.6461,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844901245640782e-05,
+      "loss": 3.6483,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84406265088973e-05,
+      "loss": 3.6389,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843227331899424e-05,
+      "loss": 3.6327,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842388737148372e-05,
+      "loss": 3.6431,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84155014239732e-05,
+      "loss": 3.632,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840711547646268e-05,
+      "loss": 3.6569,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839874590775589e-05,
+      "loss": 3.6487,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839035996024537e-05,
+      "loss": 3.639,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838197401273485e-05,
+      "loss": 3.6477,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837358806522433e-05,
+      "loss": 3.638,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836520211771381e-05,
+      "loss": 3.6457,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835683254900702e-05,
+      "loss": 3.6432,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83484466014965e-05,
+      "loss": 3.6094,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834006065398598e-05,
+      "loss": 3.6544,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833167470647546e-05,
+      "loss": 3.6409,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832328875896494e-05,
+      "loss": 3.6477,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8314919190258155e-05,
+      "loss": 3.6377,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8306533242747635e-05,
+      "loss": 3.6385,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8298147295237115e-05,
+      "loss": 3.6263,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8289761347726595e-05,
+      "loss": 3.6416,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828137540021607e-05,
+      "loss": 3.6459,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827298945270555e-05,
+      "loss": 3.6431,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826460350519503e-05,
+      "loss": 3.6557,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8256233936488244e-05,
+      "loss": 3.634,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824784798897772e-05,
+      "loss": 3.6253,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82394620414672e-05,
+      "loss": 3.6527,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823109247276041e-05,
+      "loss": 3.6301,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822270652524989e-05,
+      "loss": 3.6261,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821432057773937e-05,
+      "loss": 3.6427,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820593463022885e-05,
+      "loss": 3.6413,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819754868271833e-05,
+      "loss": 3.6263,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818917911401154e-05,
+      "loss": 3.6294,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818079316650102e-05,
+      "loss": 3.6247,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81724072189905e-05,
+      "loss": 3.6299,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816402127147998e-05,
+      "loss": 3.6445,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815563532396946e-05,
+      "loss": 3.6353,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814724937645894e-05,
+      "loss": 3.6378,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813887980775215e-05,
+      "loss": 3.6293,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813049386024163e-05,
+      "loss": 3.6549,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812210791273111e-05,
+      "loss": 3.6309,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81137219652206e-05,
+      "loss": 3.6444,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810533601771008e-05,
+      "loss": 3.6332,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809696644900329e-05,
+      "loss": 3.631,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808858050149277e-05,
+      "loss": 3.6407,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808019455398225e-05,
+      "loss": 3.6464,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071808606471727e-05,
+      "loss": 3.6352,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8063422658961207e-05,
+      "loss": 3.6404,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8055036711450686e-05,
+      "loss": 3.6305,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8046667142743896e-05,
+      "loss": 3.6274,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8038281195233376e-05,
+      "loss": 3.6362,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029895247722855e-05,
+      "loss": 3.6391,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8021509300212335e-05,
+      "loss": 3.6305,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8013123352701815e-05,
+      "loss": 3.6321,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004737405191295e-05,
+      "loss": 3.6283,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799635145768078e-05,
+      "loss": 3.6372,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798798188897399e-05,
+      "loss": 3.6386,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797959594146347e-05,
+      "loss": 3.6261,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797120999395295e-05,
+      "loss": 3.6405,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796282404644243e-05,
+      "loss": 3.6386,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7954438098931904e-05,
+      "loss": 3.6396,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794606853022512e-05,
+      "loss": 3.6167,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79376825827146e-05,
+      "loss": 3.6295,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792929663520407e-05,
+      "loss": 3.6282,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792091068769355e-05,
+      "loss": 3.6417,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791252474018303e-05,
+      "loss": 3.6369,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790413879267252e-05,
+      "loss": 3.6391,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789576922396573e-05,
+      "loss": 3.631,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788738327645521e-05,
+      "loss": 3.6296,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787899732894469e-05,
+      "loss": 3.6251,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787061138143417e-05,
+      "loss": 3.6337,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786222543392365e-05,
+      "loss": 3.6251,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785383948641313e-05,
+      "loss": 3.6351,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784546991770634e-05,
+      "loss": 3.6214,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783708397019582e-05,
+      "loss": 3.6192,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78286980226853e-05,
+      "loss": 3.6282,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782031207517478e-05,
+      "loss": 3.6291,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781192612766426e-05,
+      "loss": 3.6277,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780354018015374e-05,
+      "loss": 3.623,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779515423264322e-05,
+      "loss": 3.6369,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7786784663936433e-05,
+      "loss": 3.641,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7778398716425913e-05,
+      "loss": 3.6207,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777001276891539e-05,
+      "loss": 3.6313,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776162682140487e-05,
+      "loss": 3.6334,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775325725269808e-05,
+      "loss": 3.6297,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774487130518756e-05,
+      "loss": 3.6303,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773648535767704e-05,
+      "loss": 3.6239,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772809941016652e-05,
+      "loss": 3.6252,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7719713462656e-05,
+      "loss": 3.6266,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771134389394921e-05,
+      "loss": 3.6332,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770295794643869e-05,
+      "loss": 3.63,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769457199892817e-05,
+      "loss": 3.6242,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768618605141765e-05,
+      "loss": 3.6284,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767780010390714e-05,
+      "loss": 3.6142,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766943053520035e-05,
+      "loss": 3.6408,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766104458768983e-05,
+      "loss": 3.6228,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765265864017931e-05,
+      "loss": 3.6413,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764427269266879e-05,
+      "loss": 3.6192,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7635903123961996e-05,
+      "loss": 3.6363,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7627517176451476e-05,
+      "loss": 3.6353,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7619131228940956e-05,
+      "loss": 3.6221,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7610745281430436e-05,
+      "loss": 3.6268,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760235933391991e-05,
+      "loss": 3.6231,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759397338640939e-05,
+      "loss": 3.6371,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7585587438898876e-05,
+      "loss": 3.6439,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7577201491388356e-05,
+      "loss": 3.615,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7568831922681565e-05,
+      "loss": 3.634,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756046235397478e-05,
+      "loss": 3.6246,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755207640646426e-05,
+      "loss": 3.6329,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7543690458953734e-05,
+      "loss": 3.628,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7535304511443214e-05,
+      "loss": 3.6319,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526918563932694e-05,
+      "loss": 3.6304,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7518532616422174e-05,
+      "loss": 3.6304,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751016304771538e-05,
+      "loss": 3.631,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8950436115264893,
+      "eval_runtime": 304.1999,
+      "eval_samples_per_second": 1254.409,
+      "eval_steps_per_second": 39.201,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750177710020486e-05,
+      "loss": 3.6171,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749339115269434e-05,
+      "loss": 3.6212,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748500520518383e-05,
+      "loss": 3.6299,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747661925767331e-05,
+      "loss": 3.6294,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746823331016279e-05,
+      "loss": 3.636,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745984736265227e-05,
+      "loss": 3.6212,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745147779394548e-05,
+      "loss": 3.6232,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744309184643496e-05,
+      "loss": 3.6226,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743470589892444e-05,
+      "loss": 3.6215,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742631995141392e-05,
+      "loss": 3.6335,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74179340039034e-05,
+      "loss": 3.6216,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740954805639288e-05,
+      "loss": 3.6239,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740117848768609e-05,
+      "loss": 3.6292,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739279254017557e-05,
+      "loss": 3.6154,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738440659266505e-05,
+      "loss": 3.6232,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737602064515453e-05,
+      "loss": 3.6096,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7367634697644014e-05,
+      "loss": 3.6154,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7359248750133494e-05,
+      "loss": 3.6202,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73508791814267e-05,
+      "loss": 3.6125,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734249323391618e-05,
+      "loss": 3.6371,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733410728640566e-05,
+      "loss": 3.6342,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732572133889514e-05,
+      "loss": 3.6261,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731733539138462e-05,
+      "loss": 3.6229,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7308949443874096e-05,
+      "loss": 3.6246,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7300563496363576e-05,
+      "loss": 3.6104,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729219392765679e-05,
+      "loss": 3.6236,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728380798014627e-05,
+      "loss": 3.6228,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727542203263575e-05,
+      "loss": 3.6058,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726703608512523e-05,
+      "loss": 3.6154,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725865013761471e-05,
+      "loss": 3.6113,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.725028056890793e-05,
+      "loss": 3.6233,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72418946213974e-05,
+      "loss": 3.6185,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723350867388688e-05,
+      "loss": 3.6215,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722512272637636e-05,
+      "loss": 3.6246,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721673677886584e-05,
+      "loss": 3.614,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720835083135532e-05,
+      "loss": 3.628,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719998126264853e-05,
+      "loss": 3.6275,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719159531513801e-05,
+      "loss": 3.6133,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718320936762749e-05,
+      "loss": 3.6157,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717482342011697e-05,
+      "loss": 3.6165,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716643747260645e-05,
+      "loss": 3.6106,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7158067903899665e-05,
+      "loss": 3.6297,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7149681956389145e-05,
+      "loss": 3.6236,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7141296008878625e-05,
+      "loss": 3.6176,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132910061368105e-05,
+      "loss": 3.6226,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7124524113857585e-05,
+      "loss": 3.6171,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7116154545150794e-05,
+      "loss": 3.6197,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7107768597640274e-05,
+      "loss": 3.6219,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7099382650129754e-05,
+      "loss": 3.5888,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7090996702619234e-05,
+      "loss": 3.6288,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708262713391244e-05,
+      "loss": 3.615,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707424118640192e-05,
+      "loss": 3.6241,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70658552388914e-05,
+      "loss": 3.6183,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705746929138088e-05,
+      "loss": 3.613,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704908334387037e-05,
+      "loss": 3.606,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704071377516358e-05,
+      "loss": 3.6205,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703232782765306e-05,
+      "loss": 3.6217,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702394188014254e-05,
+      "loss": 3.6184,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701555593263202e-05,
+      "loss": 3.6319,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70071699851215e-05,
+      "loss": 3.6174,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699880041641471e-05,
+      "loss": 3.5987,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699041446890419e-05,
+      "loss": 3.6278,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698202852139367e-05,
+      "loss": 3.608,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697364257388315e-05,
+      "loss": 3.6074,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696525662637263e-05,
+      "loss": 3.6184,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695688705766584e-05,
+      "loss": 3.6186,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694850111015532e-05,
+      "loss": 3.6041,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69401151626448e-05,
+      "loss": 3.6085,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693172921513428e-05,
+      "loss": 3.6037,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692335964642749e-05,
+      "loss": 3.6078,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691497369891697e-05,
+      "loss": 3.6215,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690658775140645e-05,
+      "loss": 3.6131,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689820180389593e-05,
+      "loss": 3.6157,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889815856385405e-05,
+      "loss": 3.6067,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6881429908874885e-05,
+      "loss": 3.63,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68730603401681e-05,
+      "loss": 3.6104,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686467439265758e-05,
+      "loss": 3.6233,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685628844514706e-05,
+      "loss": 3.6065,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684790249763654e-05,
+      "loss": 3.6104,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683953292892976e-05,
+      "loss": 3.6206,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683114698141923e-05,
+      "loss": 3.6215,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682276103390871e-05,
+      "loss": 3.6099,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681437508639819e-05,
+      "loss": 3.6197,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6806005517691406e-05,
+      "loss": 3.6017,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679761957018088e-05,
+      "loss": 3.607,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678923362267036e-05,
+      "loss": 3.6148,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678084767515984e-05,
+      "loss": 3.6185,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6772478106453055e-05,
+      "loss": 3.607,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676409215894253e-05,
+      "loss": 3.6079,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6755706211432015e-05,
+      "loss": 3.6066,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6747320263921495e-05,
+      "loss": 3.6136,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738950695214704e-05,
+      "loss": 3.6165,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6730564747704184e-05,
+      "loss": 3.6051,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6722178800193664e-05,
+      "loss": 3.6162,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713792852683144e-05,
+      "loss": 3.6174,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6705406905172624e-05,
+      "loss": 3.6197,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669703733646583e-05,
+      "loss": 3.5917,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668865138895531e-05,
+      "loss": 3.6104,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668026544144479e-05,
+      "loss": 3.6055,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667187949393427e-05,
+      "loss": 3.6173,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666349354642375e-05,
+      "loss": 3.6169,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665510759891323e-05,
+      "loss": 3.6134,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664673803020645e-05,
+      "loss": 3.6057,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663835208269593e-05,
+      "loss": 3.6116,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662996613518541e-05,
+      "loss": 3.6024,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662158018767489e-05,
+      "loss": 3.6114,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66132106189681e-05,
+      "loss": 3.6054,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660482467145758e-05,
+      "loss": 3.615,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659643872394706e-05,
+      "loss": 3.602,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658805277643654e-05,
+      "loss": 3.5968,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6579683207729746e-05,
+      "loss": 3.602,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6571297260219226e-05,
+      "loss": 3.6045,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6562911312708706e-05,
+      "loss": 3.6082,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6554525365198186e-05,
+      "loss": 3.6026,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6546139417687666e-05,
+      "loss": 3.6098,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653776984898088e-05,
+      "loss": 3.6125,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652938390147036e-05,
+      "loss": 3.6046,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652099795395984e-05,
+      "loss": 3.6106,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651261200644932e-05,
+      "loss": 3.6102,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650424243774253e-05,
+      "loss": 3.6086,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649585649023201e-05,
+      "loss": 3.6049,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648747054272149e-05,
+      "loss": 3.6015,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647908459521097e-05,
+      "loss": 3.6047,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647071502650418e-05,
+      "loss": 3.603,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646232907899366e-05,
+      "loss": 3.6119,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645394313148314e-05,
+      "loss": 3.6112,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644555718397262e-05,
+      "loss": 3.603,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64371712364621e-05,
+      "loss": 3.6036,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6428801667755316e-05,
+      "loss": 3.5942,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6420415720244796e-05,
+      "loss": 3.6177,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6412029772734276e-05,
+      "loss": 3.6003,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6403643825223756e-05,
+      "loss": 3.6165,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6395257877713236e-05,
+      "loss": 3.6004,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386888309006445e-05,
+      "loss": 3.613,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6378502361495925e-05,
+      "loss": 3.6133,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6370116413985405e-05,
+      "loss": 3.5989,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6361730466474884e-05,
+      "loss": 3.6092,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6353360897768094e-05,
+      "loss": 3.5987,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6344974950257574e-05,
+      "loss": 3.6171,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6336589002747053e-05,
+      "loss": 3.6163,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632820305523654e-05,
+      "loss": 3.5997,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319817107726013e-05,
+      "loss": 3.6067,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631144753901923e-05,
+      "loss": 3.6038,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630306159150871e-05,
+      "loss": 3.6134,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629467564399819e-05,
+      "loss": 3.6029,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628628969648766e-05,
+      "loss": 3.6143,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627790374897714e-05,
+      "loss": 3.6036,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626953418027036e-05,
+      "loss": 3.6126,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626114823275983e-05,
+      "loss": 3.6044,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.893688440322876,
+      "eval_runtime": 302.4087,
+      "eval_samples_per_second": 1261.839,
+      "eval_steps_per_second": 39.433,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625276228524931e-05,
+      "loss": 3.6045,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624437633773879e-05,
+      "loss": 3.5992,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623599039022828e-05,
+      "loss": 3.6049,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622762082152149e-05,
+      "loss": 3.6111,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621923487401097e-05,
+      "loss": 3.6157,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621084892650045e-05,
+      "loss": 3.599,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620246297898993e-05,
+      "loss": 3.6007,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6194093410283136e-05,
+      "loss": 3.6029,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6185707462772616e-05,
+      "loss": 3.6044,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6177321515262096e-05,
+      "loss": 3.6061,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168935567751576e-05,
+      "loss": 3.5982,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6160565999044785e-05,
+      "loss": 3.6017,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6152196430338e-05,
+      "loss": 3.6105,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614381048282748e-05,
+      "loss": 3.5949,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613542453531696e-05,
+      "loss": 3.5977,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612703858780644e-05,
+      "loss": 3.5886,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611865264029592e-05,
+      "loss": 3.5905,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61102666927854e-05,
+      "loss": 3.6,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610188074527488e-05,
+      "loss": 3.5938,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609349479776436e-05,
+      "loss": 3.6112,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608512522905757e-05,
+      "loss": 3.611,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607673928154705e-05,
+      "loss": 3.6043,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606835333403653e-05,
+      "loss": 3.6028,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605996738652601e-05,
+      "loss": 3.6015,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605158143901549e-05,
+      "loss": 3.5876,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60432118703087e-05,
+      "loss": 3.6094,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6034825922798185e-05,
+      "loss": 3.5996,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6026439975287665e-05,
+      "loss": 3.5868,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6018054027777145e-05,
+      "loss": 3.5942,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6009684459070354e-05,
+      "loss": 3.5935,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6001298511559834e-05,
+      "loss": 3.5985,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5992912564049314e-05,
+      "loss": 3.5952,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5984526616538794e-05,
+      "loss": 3.605,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5976140669028274e-05,
+      "loss": 3.6021,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5967754721517754e-05,
+      "loss": 3.5918,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595938515281096e-05,
+      "loss": 3.6076,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595099920530044e-05,
+      "loss": 3.6075,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594261325778992e-05,
+      "loss": 3.5926,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59342273102794e-05,
+      "loss": 3.5968,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592585774157262e-05,
+      "loss": 3.5957,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59174717940621e-05,
+      "loss": 3.5864,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590908584655158e-05,
+      "loss": 3.61,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590069989904106e-05,
+      "loss": 3.6023,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589233033033427e-05,
+      "loss": 3.6023,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588394438282375e-05,
+      "loss": 3.5974,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587555843531323e-05,
+      "loss": 3.6003,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586717248780271e-05,
+      "loss": 3.599,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585880291909592e-05,
+      "loss": 3.5985,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58504169715854e-05,
+      "loss": 3.5712,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584203102407488e-05,
+      "loss": 3.6083,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583364507656436e-05,
+      "loss": 3.593,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582525912905384e-05,
+      "loss": 3.6056,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581688956034705e-05,
+      "loss": 3.596,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580850361283653e-05,
+      "loss": 3.5916,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580011766532601e-05,
+      "loss": 3.5828,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579173171781549e-05,
+      "loss": 3.5953,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57833621491087e-05,
+      "loss": 3.6076,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577497620159818e-05,
+      "loss": 3.5955,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576659025408766e-05,
+      "loss": 3.6115,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575820430657714e-05,
+      "loss": 3.5955,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574983473787035e-05,
+      "loss": 3.5779,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574144879035983e-05,
+      "loss": 3.606,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573306284284931e-05,
+      "loss": 3.5908,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572467689533879e-05,
+      "loss": 3.5859,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571629094782827e-05,
+      "loss": 3.5943,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5707921379121486e-05,
+      "loss": 3.5996,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5699535431610966e-05,
+      "loss": 3.5794,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569114948410044e-05,
+      "loss": 3.5891,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568276353658992e-05,
+      "loss": 3.5818,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56743775890794e-05,
+      "loss": 3.5937,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5666008020372615e-05,
+      "loss": 3.5966,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565762207286209e-05,
+      "loss": 3.5934,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564923612535157e-05,
+      "loss": 3.5987,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564085017784105e-05,
+      "loss": 3.5849,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5632480609134264e-05,
+      "loss": 3.6063,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5624094661623744e-05,
+      "loss": 3.5944,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5615708714113224e-05,
+      "loss": 3.6062,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5607322766602704e-05,
+      "loss": 3.5873,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559895319789591e-05,
+      "loss": 3.5896,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559056725038539e-05,
+      "loss": 3.5985,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558218130287487e-05,
+      "loss": 3.5994,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557379535536435e-05,
+      "loss": 3.5957,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556540940785383e-05,
+      "loss": 3.5989,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555702346034331e-05,
+      "loss": 3.5834,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554865389163652e-05,
+      "loss": 3.5878,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5540267944126e-05,
+      "loss": 3.592,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553188199661548e-05,
+      "loss": 3.5992,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552349604910496e-05,
+      "loss": 3.5844,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551512648039818e-05,
+      "loss": 3.5889,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550674053288766e-05,
+      "loss": 3.5888,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549835458537714e-05,
+      "loss": 3.5899,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548996863786662e-05,
+      "loss": 3.5951,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548159906915983e-05,
+      "loss": 3.5841,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5473213121649307e-05,
+      "loss": 3.5933,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5464827174138787e-05,
+      "loss": 3.5998,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5456441226628266e-05,
+      "loss": 3.5959,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5448071657921476e-05,
+      "loss": 3.5753,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5439685710410956e-05,
+      "loss": 3.5862,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5431299762900436e-05,
+      "loss": 3.5838,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5422913815389915e-05,
+      "loss": 3.5985,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54145278678794e-05,
+      "loss": 3.5965,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540614192036888e-05,
+      "loss": 3.5944,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539777235166209e-05,
+      "loss": 3.5844,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538938640415157e-05,
+      "loss": 3.5907,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538100045664105e-05,
+      "loss": 3.5836,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537261450913053e-05,
+      "loss": 3.5905,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536424494042374e-05,
+      "loss": 3.5861,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535585899291322e-05,
+      "loss": 3.5924,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53474730454027e-05,
+      "loss": 3.5856,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533908709789218e-05,
+      "loss": 3.5765,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533071752918539e-05,
+      "loss": 3.5822,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532233158167487e-05,
+      "loss": 3.5824,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313945634164356e-05,
+      "loss": 3.5885,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5305559686653836e-05,
+      "loss": 3.5833,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5297190117947045e-05,
+      "loss": 3.592,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5288804170436525e-05,
+      "loss": 3.5906,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5280418222926005e-05,
+      "loss": 3.5884,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5272032275415485e-05,
+      "loss": 3.5898,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5263662706708694e-05,
+      "loss": 3.5885,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5255276759198174e-05,
+      "loss": 3.5913,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5246890811687654e-05,
+      "loss": 3.5846,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5238504864177134e-05,
+      "loss": 3.5816,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5230118916666614e-05,
+      "loss": 3.5868,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522174934795982e-05,
+      "loss": 3.5842,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521336340044931e-05,
+      "loss": 3.5914,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520497745293879e-05,
+      "loss": 3.5873,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519659150542827e-05,
+      "loss": 3.5839,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518822193672148e-05,
+      "loss": 3.5824,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517983598921096e-05,
+      "loss": 3.5781,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517145004170044e-05,
+      "loss": 3.5942,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516306409418992e-05,
+      "loss": 3.586,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515469452548313e-05,
+      "loss": 3.5986,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514630857797261e-05,
+      "loss": 3.5795,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513792263046209e-05,
+      "loss": 3.5904,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512953668295157e-05,
+      "loss": 3.5932,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5121167114244776e-05,
+      "loss": 3.5832,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511278116673426e-05,
+      "loss": 3.5823,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510439521922374e-05,
+      "loss": 3.583,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509600927171322e-05,
+      "loss": 3.5949,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508763970300643e-05,
+      "loss": 3.5965,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507925375549591e-05,
+      "loss": 3.5838,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507086780798539e-05,
+      "loss": 3.582,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506248186047487e-05,
+      "loss": 3.5879,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5054095912964345e-05,
+      "loss": 3.5917,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504572634425756e-05,
+      "loss": 3.5825,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503734039674704e-05,
+      "loss": 3.5951,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028954449236514e-05,
+      "loss": 3.5822,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5020568501726e-05,
+      "loss": 3.5942,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501219893301922e-05,
+      "loss": 3.585,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.894435405731201,
+      "eval_runtime": 302.4864,
+      "eval_samples_per_second": 1261.514,
+      "eval_steps_per_second": 39.423,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.50038129855087e-05,
+      "loss": 3.59,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499542703799817e-05,
+      "loss": 3.5752,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498704109048765e-05,
+      "loss": 3.5846,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497865514297713e-05,
+      "loss": 3.5914,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497026919546661e-05,
+      "loss": 3.5934,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496189962675982e-05,
+      "loss": 3.5846,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49535136792493e-05,
+      "loss": 3.582,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494512773173878e-05,
+      "loss": 3.5829,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493674178422826e-05,
+      "loss": 3.5823,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492837221552147e-05,
+      "loss": 3.5847,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4919986268010955e-05,
+      "loss": 3.5841,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4911600320500435e-05,
+      "loss": 3.5813,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4903230751793644e-05,
+      "loss": 3.5935,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4894844804283124e-05,
+      "loss": 3.5733,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4886458856772604e-05,
+      "loss": 3.5779,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4878072909262084e-05,
+      "loss": 3.5695,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4869686961751564e-05,
+      "loss": 3.5708,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4861301014241043e-05,
+      "loss": 3.583,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4852915066730523e-05,
+      "loss": 3.5739,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4844529119220003e-05,
+      "loss": 3.5918,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483615955051321e-05,
+      "loss": 3.5924,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482777360300269e-05,
+      "loss": 3.5857,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481938765549217e-05,
+      "loss": 3.5829,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481100170798165e-05,
+      "loss": 3.5843,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480261576047113e-05,
+      "loss": 3.5673,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479424619176435e-05,
+      "loss": 3.593,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478586024425383e-05,
+      "loss": 3.5768,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477747429674331e-05,
+      "loss": 3.571,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476908834923279e-05,
+      "loss": 3.5787,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4760718780526e-05,
+      "loss": 3.5724,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475233283301548e-05,
+      "loss": 3.5791,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474394688550496e-05,
+      "loss": 3.5754,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473556093799444e-05,
+      "loss": 3.5915,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472717499048392e-05,
+      "loss": 3.5789,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4718805421777126e-05,
+      "loss": 3.576,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4710419474266606e-05,
+      "loss": 3.5852,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4702033526756086e-05,
+      "loss": 3.589,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469364757924557e-05,
+      "loss": 3.5739,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468527801053878e-05,
+      "loss": 3.5734,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467689206302826e-05,
+      "loss": 3.5767,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466850611551774e-05,
+      "loss": 3.5709,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466012016800722e-05,
+      "loss": 3.5855,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465175059930043e-05,
+      "loss": 3.5825,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464336465178991e-05,
+      "loss": 3.586,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463497870427939e-05,
+      "loss": 3.58,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462659275676887e-05,
+      "loss": 3.5769,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461822318806208e-05,
+      "loss": 3.5801,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460983724055156e-05,
+      "loss": 3.579,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460145129304104e-05,
+      "loss": 3.553,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4593065345530526e-05,
+      "loss": 3.5862,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4584695776823735e-05,
+      "loss": 3.5763,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4576309829313215e-05,
+      "loss": 3.5854,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4567923881802695e-05,
+      "loss": 3.5752,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4559537934292175e-05,
+      "loss": 3.5784,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4551168365585384e-05,
+      "loss": 3.5626,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4542782418074864e-05,
+      "loss": 3.5725,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4534396470564344e-05,
+      "loss": 3.5908,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4526010523053824e-05,
+      "loss": 3.5764,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4517640954347033e-05,
+      "loss": 3.5909,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450925500683651e-05,
+      "loss": 3.5782,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450086905932599e-05,
+      "loss": 3.557,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449248311181548e-05,
+      "loss": 3.591,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448411354310869e-05,
+      "loss": 3.57,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447572759559817e-05,
+      "loss": 3.5668,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446734164808765e-05,
+      "loss": 3.5789,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445895570057712e-05,
+      "loss": 3.5772,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445058613187034e-05,
+      "loss": 3.5647,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444220018435982e-05,
+      "loss": 3.569,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44338142368493e-05,
+      "loss": 3.565,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442542828933877e-05,
+      "loss": 3.5677,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441704234182825e-05,
+      "loss": 3.5841,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440867277312147e-05,
+      "loss": 3.5747,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440028682561095e-05,
+      "loss": 3.5766,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439190087810043e-05,
+      "loss": 3.5666,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438351493058991e-05,
+      "loss": 3.5882,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437514536188312e-05,
+      "loss": 3.5747,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4366759414372596e-05,
+      "loss": 3.5859,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4358373466862076e-05,
+      "loss": 3.5666,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4349987519351556e-05,
+      "loss": 3.5709,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434161795064477e-05,
+      "loss": 3.5826,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4333232003134245e-05,
+      "loss": 3.5797,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4324846055623725e-05,
+      "loss": 3.5779,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4316460108113205e-05,
+      "loss": 3.5812,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4308074160602685e-05,
+      "loss": 3.5649,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42997045918959e-05,
+      "loss": 3.5665,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429131864438538e-05,
+      "loss": 3.5745,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428293269687486e-05,
+      "loss": 3.5788,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427454674936434e-05,
+      "loss": 3.5687,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426616080185382e-05,
+      "loss": 3.5742,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425779123314703e-05,
+      "loss": 3.5684,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424940528563651e-05,
+      "loss": 3.5723,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424101933812599e-05,
+      "loss": 3.5754,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423263339061547e-05,
+      "loss": 3.5692,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422426382190868e-05,
+      "loss": 3.5752,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421587787439816e-05,
+      "loss": 3.5821,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420749192688764e-05,
+      "loss": 3.5814,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4199105979377125e-05,
+      "loss": 3.5521,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4190736410670334e-05,
+      "loss": 3.5702,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4182350463159814e-05,
+      "loss": 3.565,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4173964515649294e-05,
+      "loss": 3.5794,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4165578568138774e-05,
+      "loss": 3.5793,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4157192620628254e-05,
+      "loss": 3.5761,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414882305192146e-05,
+      "loss": 3.5648,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414043710441094e-05,
+      "loss": 3.5727,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413205115690042e-05,
+      "loss": 3.5677,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41236652093899e-05,
+      "loss": 3.5695,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411529564068311e-05,
+      "loss": 3.566,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410690969317259e-05,
+      "loss": 3.5751,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409852374566207e-05,
+      "loss": 3.567,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409013779815156e-05,
+      "loss": 3.56,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408176822944477e-05,
+      "loss": 3.5666,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407338228193425e-05,
+      "loss": 3.5613,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406499633442373e-05,
+      "loss": 3.5725,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405661038691321e-05,
+      "loss": 3.5635,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404824081820642e-05,
+      "loss": 3.5736,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40398548706959e-05,
+      "loss": 3.5736,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403146892318538e-05,
+      "loss": 3.5712,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402308297567486e-05,
+      "loss": 3.5685,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4014713406968066e-05,
+      "loss": 3.5739,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4006327459457546e-05,
+      "loss": 3.5731,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3997941511947026e-05,
+      "loss": 3.5647,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398955556443651e-05,
+      "loss": 3.5625,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398116961692599e-05,
+      "loss": 3.5691,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39728000482192e-05,
+      "loss": 3.5645,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396441410070868e-05,
+      "loss": 3.5728,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395602815319816e-05,
+      "loss": 3.5682,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394764220568764e-05,
+      "loss": 3.5673,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393927263698085e-05,
+      "loss": 3.5659,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393088668947033e-05,
+      "loss": 3.5592,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392250074195981e-05,
+      "loss": 3.5724,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391411479444929e-05,
+      "loss": 3.5696,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39057452257425e-05,
+      "loss": 3.5819,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389735927823198e-05,
+      "loss": 3.561,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888973330721466e-05,
+      "loss": 3.5685,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3880587383210946e-05,
+      "loss": 3.5775,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3872217814504155e-05,
+      "loss": 3.5681,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3863831866993635e-05,
+      "loss": 3.5611,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3855445919483115e-05,
+      "loss": 3.5648,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3847059971972595e-05,
+      "loss": 3.5784,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3838674024462075e-05,
+      "loss": 3.5784,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3830304455755284e-05,
+      "loss": 3.5701,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3821918508244764e-05,
+      "loss": 3.5645,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3813532560734244e-05,
+      "loss": 3.5695,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3805146613223724e-05,
+      "loss": 3.5713,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379677704451693e-05,
+      "loss": 3.561,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378839109700642e-05,
+      "loss": 3.585,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37800051494959e-05,
+      "loss": 3.56,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377161920198538e-05,
+      "loss": 3.5755,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376324963327859e-05,
+      "loss": 3.5657,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.894693374633789,
+      "eval_runtime": 304.9288,
+      "eval_samples_per_second": 1251.41,
+      "eval_steps_per_second": 39.107,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375486368576807e-05,
+      "loss": 3.5729,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374647773825755e-05,
+      "loss": 3.5578,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373809179074703e-05,
+      "loss": 3.5646,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372972222204024e-05,
+      "loss": 3.5769,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372133627452972e-05,
+      "loss": 3.5704,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.37129503270192e-05,
+      "loss": 3.5704,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370456437950867e-05,
+      "loss": 3.5641,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369617843199816e-05,
+      "loss": 3.5615,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368779248448764e-05,
+      "loss": 3.5646,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3679422915780853e-05,
+      "loss": 3.5702,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3671036968270327e-05,
+      "loss": 3.5635,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3662651020759807e-05,
+      "loss": 3.57,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365428145205302e-05,
+      "loss": 3.5729,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36458955045425e-05,
+      "loss": 3.5547,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3637509557031976e-05,
+      "loss": 3.5636,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3629123609521456e-05,
+      "loss": 3.5486,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3620737662010935e-05,
+      "loss": 3.5515,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3612351714500415e-05,
+      "loss": 3.5689,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3603965766989895e-05,
+      "loss": 3.5543,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3595579819479375e-05,
+      "loss": 3.5724,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358721025077259e-05,
+      "loss": 3.5763,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357882430326207e-05,
+      "loss": 3.566,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357043835575155e-05,
+      "loss": 3.5674,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356205240824103e-05,
+      "loss": 3.5677,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355368283953424e-05,
+      "loss": 3.552,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354529689202372e-05,
+      "loss": 3.5726,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35369109445132e-05,
+      "loss": 3.5585,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352852499700268e-05,
+      "loss": 3.5558,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352013904949216e-05,
+      "loss": 3.5613,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351176948078537e-05,
+      "loss": 3.5504,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350338353327485e-05,
+      "loss": 3.5581,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349499758576433e-05,
+      "loss": 3.5613,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348661163825381e-05,
+      "loss": 3.5718,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3478242069547025e-05,
+      "loss": 3.5602,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3469856122036505e-05,
+      "loss": 3.5579,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3461470174525985e-05,
+      "loss": 3.5618,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3453084227015465e-05,
+      "loss": 3.5742,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3444698279504945e-05,
+      "loss": 3.5589,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3436328710798154e-05,
+      "loss": 3.5553,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3427942763287634e-05,
+      "loss": 3.5592,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3419556815777114e-05,
+      "loss": 3.5593,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3411170868266594e-05,
+      "loss": 3.562,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34028012995598e-05,
+      "loss": 3.5707,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339441535204928e-05,
+      "loss": 3.5654,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338602940453876e-05,
+      "loss": 3.5589,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337764345702824e-05,
+      "loss": 3.5603,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336927388832146e-05,
+      "loss": 3.5616,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336088794081094e-05,
+      "loss": 3.5621,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335250199330042e-05,
+      "loss": 3.5394,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33441160457899e-05,
+      "loss": 3.5653,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333574647708311e-05,
+      "loss": 3.5601,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332736052957259e-05,
+      "loss": 3.5667,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331897458206207e-05,
+      "loss": 3.5558,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331058863455155e-05,
+      "loss": 3.5588,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3302219065844756e-05,
+      "loss": 3.5548,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3293833118334236e-05,
+      "loss": 3.5505,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3285447170823716e-05,
+      "loss": 3.5734,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3277061223313196e-05,
+      "loss": 3.5574,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326869165460641e-05,
+      "loss": 3.5793,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326030570709589e-05,
+      "loss": 3.5582,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325191975958537e-05,
+      "loss": 3.5434,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324353381207485e-05,
+      "loss": 3.5705,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323516424336806e-05,
+      "loss": 3.5565,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322677829585754e-05,
+      "loss": 3.5464,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321839234834702e-05,
+      "loss": 3.5554,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32100064008365e-05,
+      "loss": 3.5667,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320163683212971e-05,
+      "loss": 3.5443,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319325088461919e-05,
+      "loss": 3.5536,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318486493710867e-05,
+      "loss": 3.5476,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317647898959815e-05,
+      "loss": 3.5477,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316809304208764e-05,
+      "loss": 3.5672,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3159723473380846e-05,
+      "loss": 3.5575,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3151337525870326e-05,
+      "loss": 3.5583,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3142951578359806e-05,
+      "loss": 3.5499,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313456563084928e-05,
+      "loss": 3.5691,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3126196062142495e-05,
+      "loss": 3.5595,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3117810114631975e-05,
+      "loss": 3.5673,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3109424167121455e-05,
+      "loss": 3.5515,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310103821961093e-05,
+      "loss": 3.552,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3092668650904144e-05,
+      "loss": 3.5647,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3084282703393624e-05,
+      "loss": 3.5603,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3075896755883104e-05,
+      "loss": 3.5599,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3067510808372584e-05,
+      "loss": 3.5636,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30591412396658e-05,
+      "loss": 3.5534,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305075529215528e-05,
+      "loss": 3.5451,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304236934464475e-05,
+      "loss": 3.5576,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303398339713423e-05,
+      "loss": 3.5617,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302561382842745e-05,
+      "loss": 3.5529,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301722788091693e-05,
+      "loss": 3.5558,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30088419334064e-05,
+      "loss": 3.5515,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.300045598589588e-05,
+      "loss": 3.5541,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299207003838536e-05,
+      "loss": 3.5562,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298370046967858e-05,
+      "loss": 3.5534,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297531452216806e-05,
+      "loss": 3.5591,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296692857465754e-05,
+      "loss": 3.5621,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295854262714702e-05,
+      "loss": 3.5653,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2950173058440226e-05,
+      "loss": 3.5367,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2941787110929706e-05,
+      "loss": 3.5521,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2933401163419186e-05,
+      "loss": 3.5496,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2925015215908666e-05,
+      "loss": 3.56,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2916645647201875e-05,
+      "loss": 3.5641,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2908259699691355e-05,
+      "loss": 3.5556,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2899873752180835e-05,
+      "loss": 3.5477,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2891487804670315e-05,
+      "loss": 3.5551,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288311823596353e-05,
+      "loss": 3.5521,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287473228845301e-05,
+      "loss": 3.5531,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286634634094249e-05,
+      "loss": 3.5493,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285796039343197e-05,
+      "loss": 3.5567,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284959082472518e-05,
+      "loss": 3.5516,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284120487721466e-05,
+      "loss": 3.5421,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283281892970414e-05,
+      "loss": 3.5522,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282443298219362e-05,
+      "loss": 3.5382,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281606341348683e-05,
+      "loss": 3.5551,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280767746597631e-05,
+      "loss": 3.5513,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279929151846579e-05,
+      "loss": 3.5512,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279090557095527e-05,
+      "loss": 3.5578,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2782536002248485e-05,
+      "loss": 3.554,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2774150054737965e-05,
+      "loss": 3.5496,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2765764107227445e-05,
+      "loss": 3.5537,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2757378159716925e-05,
+      "loss": 3.56,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2749008591010134e-05,
+      "loss": 3.5463,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2740622643499614e-05,
+      "loss": 3.5433,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2732236695989094e-05,
+      "loss": 3.554,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2723850748478573e-05,
+      "loss": 3.5493,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271548117977178e-05,
+      "loss": 3.5531,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270709523226126e-05,
+      "loss": 3.5517,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269870928475074e-05,
+      "loss": 3.553,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269032333724022e-05,
+      "loss": 3.5518,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268195376853344e-05,
+      "loss": 3.5434,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267356782102292e-05,
+      "loss": 3.5459,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26651818735124e-05,
+      "loss": 3.5575,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265679592600188e-05,
+      "loss": 3.5646,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264842635729509e-05,
+      "loss": 3.5435,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264004040978457e-05,
+      "loss": 3.5471,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263165446227405e-05,
+      "loss": 3.5628,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262326851476353e-05,
+      "loss": 3.552,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261488256725301e-05,
+      "loss": 3.546,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2606512998546216e-05,
+      "loss": 3.5472,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2598127051035696e-05,
+      "loss": 3.5623,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2589741103525176e-05,
+      "loss": 3.5601,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2581355156014656e-05,
+      "loss": 3.5535,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257298558730787e-05,
+      "loss": 3.5448,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256459963979735e-05,
+      "loss": 3.5528,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255621369228683e-05,
+      "loss": 3.5537,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254782774477631e-05,
+      "loss": 3.543,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253945817606952e-05,
+      "loss": 3.5675,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2531072228559e-05,
+      "loss": 3.5452,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252268628104848e-05,
+      "loss": 3.5579,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251430033353796e-05,
+      "loss": 3.5512,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.896216630935669,
+      "eval_runtime": 312.4797,
+      "eval_samples_per_second": 1221.171,
+      "eval_steps_per_second": 38.162,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250593076483117e-05,
+      "loss": 3.5756,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249754481732065e-05,
+      "loss": 3.5388,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248915886981013e-05,
+      "loss": 3.5486,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248077292229961e-05,
+      "loss": 3.559,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247238697478909e-05,
+      "loss": 3.5542,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2464017406082306e-05,
+      "loss": 3.5521,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2455631458571786e-05,
+      "loss": 3.5496,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2447245511061265e-05,
+      "loss": 3.5455,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2438859563550745e-05,
+      "loss": 3.5475,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2430489994843955e-05,
+      "loss": 3.5535,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2422104047333434e-05,
+      "loss": 3.548,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2413718099822914e-05,
+      "loss": 3.5535,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2405348531116124e-05,
+      "loss": 3.555,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2396962583605603e-05,
+      "loss": 3.5381,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2388576636095083e-05,
+      "loss": 3.5481,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238019068858456e-05,
+      "loss": 3.5329,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237180474107404e-05,
+      "loss": 3.5364,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236341879356353e-05,
+      "loss": 3.5479,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235503284605301e-05,
+      "loss": 3.5365,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234664689854248e-05,
+      "loss": 3.5535,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23382773298357e-05,
+      "loss": 3.563,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232989138232518e-05,
+      "loss": 3.5453,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232150543481466e-05,
+      "loss": 3.5527,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231311948730413e-05,
+      "loss": 3.5498,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230474991859735e-05,
+      "loss": 3.5359,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229636397108683e-05,
+      "loss": 3.5539,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22879780235763e-05,
+      "loss": 3.5401,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227959207606578e-05,
+      "loss": 3.5431,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227120612855527e-05,
+      "loss": 3.5409,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2262836559848484e-05,
+      "loss": 3.5337,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225445061233796e-05,
+      "loss": 3.5408,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224606466482744e-05,
+      "loss": 3.5475,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223767871731692e-05,
+      "loss": 3.5535,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222930914861013e-05,
+      "loss": 3.5446,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2220923201099606e-05,
+      "loss": 3.5386,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2212537253589086e-05,
+      "loss": 3.5449,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2204151306078566e-05,
+      "loss": 3.5603,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2195765358568046e-05,
+      "loss": 3.5397,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2187395789861255e-05,
+      "loss": 3.5428,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2179009842350735e-05,
+      "loss": 3.543,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217062389484022e-05,
+      "loss": 3.5386,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21622379473297e-05,
+      "loss": 3.5463,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215386837862291e-05,
+      "loss": 3.5525,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214548243111239e-05,
+      "loss": 3.553,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213709648360187e-05,
+      "loss": 3.5424,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212871053609135e-05,
+      "loss": 3.5433,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212034096738456e-05,
+      "loss": 3.542,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211195501987404e-05,
+      "loss": 3.5512,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210356907236352e-05,
+      "loss": 3.5208,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2095183124853e-05,
+      "loss": 3.5476,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208681355614621e-05,
+      "loss": 3.5475,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207842760863569e-05,
+      "loss": 3.5487,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2070041661125175e-05,
+      "loss": 3.5414,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2061655713614655e-05,
+      "loss": 3.5364,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2053286144907864e-05,
+      "loss": 3.5423,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2044900197397344e-05,
+      "loss": 3.533,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2036514249886824e-05,
+      "loss": 3.5551,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2028128302376304e-05,
+      "loss": 3.5431,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201975873366951e-05,
+      "loss": 3.5591,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201137278615899e-05,
+      "loss": 3.5411,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200298683864847e-05,
+      "loss": 3.5265,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199460089113795e-05,
+      "loss": 3.5562,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198623132243116e-05,
+      "loss": 3.5369,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197784537492064e-05,
+      "loss": 3.5318,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196945942741013e-05,
+      "loss": 3.542,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196107347989961e-05,
+      "loss": 3.5471,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195270391119282e-05,
+      "loss": 3.5282,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19443179636823e-05,
+      "loss": 3.537,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193593201617178e-05,
+      "loss": 3.5332,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192754606866126e-05,
+      "loss": 3.532,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191916012115074e-05,
+      "loss": 3.549,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191079055244395e-05,
+      "loss": 3.5429,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190240460493343e-05,
+      "loss": 3.5401,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189401865742291e-05,
+      "loss": 3.5382,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188563270991239e-05,
+      "loss": 3.5503,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1877263141205596e-05,
+      "loss": 3.5414,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186887719369508e-05,
+      "loss": 3.5489,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186049124618456e-05,
+      "loss": 3.5403,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185210529867404e-05,
+      "loss": 3.5367,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184373572996725e-05,
+      "loss": 3.5465,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183534978245673e-05,
+      "loss": 3.545,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182696383494621e-05,
+      "loss": 3.5451,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181857788743569e-05,
+      "loss": 3.5476,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18102083187289e-05,
+      "loss": 3.5434,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180182237121838e-05,
+      "loss": 3.5256,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179343642370786e-05,
+      "loss": 3.5404,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178505047619734e-05,
+      "loss": 3.5496,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177668090749055e-05,
+      "loss": 3.5367,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176829495998003e-05,
+      "loss": 3.5392,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1759909012469516e-05,
+      "loss": 3.5356,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1751523064958996e-05,
+      "loss": 3.538,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1743137117448476e-05,
+      "loss": 3.5368,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1734767548741685e-05,
+      "loss": 3.5399,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1726381601231165e-05,
+      "loss": 3.545,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1717995653720645e-05,
+      "loss": 3.543,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1709609706210125e-05,
+      "loss": 3.5494,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1701240137503334e-05,
+      "loss": 3.5188,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1692854189992814e-05,
+      "loss": 3.5395,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1684468242482294e-05,
+      "loss": 3.534,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1676082294971774e-05,
+      "loss": 3.5434,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166771272626498e-05,
+      "loss": 3.5443,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165932677875447e-05,
+      "loss": 3.543,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165094083124395e-05,
+      "loss": 3.532,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164255488373343e-05,
+      "loss": 3.543,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163418531502664e-05,
+      "loss": 3.5401,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162579936751612e-05,
+      "loss": 3.5326,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16174134200056e-05,
+      "loss": 3.5331,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160902747249508e-05,
+      "loss": 3.538,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160065790378829e-05,
+      "loss": 3.5413,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159227195627777e-05,
+      "loss": 3.5258,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158388600876725e-05,
+      "loss": 3.5378,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157550006125673e-05,
+      "loss": 3.5235,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156713049254994e-05,
+      "loss": 3.5426,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1558744545039423e-05,
+      "loss": 3.5332,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1550358597528903e-05,
+      "loss": 3.5348,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1541972650018383e-05,
+      "loss": 3.5399,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153360308131159e-05,
+      "loss": 3.5408,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152521713380107e-05,
+      "loss": 3.5331,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151683118629055e-05,
+      "loss": 3.5399,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150844523878003e-05,
+      "loss": 3.5428,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150007567007324e-05,
+      "loss": 3.528,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149168972256272e-05,
+      "loss": 3.5267,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14833037750522e-05,
+      "loss": 3.5405,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147491782754168e-05,
+      "loss": 3.5383,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146654825883489e-05,
+      "loss": 3.5392,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145816231132438e-05,
+      "loss": 3.5336,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144977636381386e-05,
+      "loss": 3.5358,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144139041630333e-05,
+      "loss": 3.5338,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1433020847596546e-05,
+      "loss": 3.5267,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1424634900086026e-05,
+      "loss": 3.5289,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1416248952575506e-05,
+      "loss": 3.5446,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140786300506498e-05,
+      "loss": 3.548,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1399493436358195e-05,
+      "loss": 3.532,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1391107488847675e-05,
+      "loss": 3.5319,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1382721541337155e-05,
+      "loss": 3.5429,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137433559382663e-05,
+      "loss": 3.5355,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1365966025119844e-05,
+      "loss": 3.5298,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135758007760933e-05,
+      "loss": 3.5362,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1349194130098804e-05,
+      "loss": 3.5471,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1340808182588284e-05,
+      "loss": 3.5421,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13324386138815e-05,
+      "loss": 3.5376,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132405266637098e-05,
+      "loss": 3.5278,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131566671886045e-05,
+      "loss": 3.5398,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130728077134993e-05,
+      "loss": 3.5394,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129891120264315e-05,
+      "loss": 3.5266,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129052525513262e-05,
+      "loss": 3.5532,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12821393076221e-05,
+      "loss": 3.532,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127375336011158e-05,
+      "loss": 3.5416,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12653837914048e-05,
+      "loss": 3.5323,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.89788556098938,
+      "eval_runtime": 303.7745,
+      "eval_samples_per_second": 1256.165,
+      "eval_steps_per_second": 39.256,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125699784389428e-05,
+      "loss": 3.5387,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124861189638376e-05,
+      "loss": 3.5288,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124022594887324e-05,
+      "loss": 3.5342,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123184000136272e-05,
+      "loss": 3.5416,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122347043265593e-05,
+      "loss": 3.5402,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121508448514541e-05,
+      "loss": 3.5381,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1206698537634887e-05,
+      "loss": 3.5292,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1198328968928096e-05,
+      "loss": 3.5318,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1189943021417576e-05,
+      "loss": 3.5319,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1181557073907056e-05,
+      "loss": 3.5377,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1173171126396536e-05,
+      "loss": 3.5322,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116480155768975e-05,
+      "loss": 3.5354,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115643198898297e-05,
+      "loss": 3.5428,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114804604147245e-05,
+      "loss": 3.525,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113966009396193e-05,
+      "loss": 3.5306,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11312741464514e-05,
+      "loss": 3.5148,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112288819894088e-05,
+      "loss": 3.5227,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111450225143036e-05,
+      "loss": 3.531,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110611630391984e-05,
+      "loss": 3.5236,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109773035640932e-05,
+      "loss": 3.5338,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108936078770253e-05,
+      "loss": 3.5499,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108097484019201e-05,
+      "loss": 3.5338,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107258889268149e-05,
+      "loss": 3.5314,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1064202945170976e-05,
+      "loss": 3.5346,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1055833376464185e-05,
+      "loss": 3.5274,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1047447428953665e-05,
+      "loss": 3.5325,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1039061481443145e-05,
+      "loss": 3.5224,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1030675533932625e-05,
+      "loss": 3.5328,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1022289586422105e-05,
+      "loss": 3.5277,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1013920017715314e-05,
+      "loss": 3.5175,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1005534070204794e-05,
+      "loss": 3.5288,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0997148122694274e-05,
+      "loss": 3.5257,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0988762175183754e-05,
+      "loss": 3.5364,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098039260647696e-05,
+      "loss": 3.5338,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097200665896644e-05,
+      "loss": 3.5232,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096362071145593e-05,
+      "loss": 3.5276,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095523476394541e-05,
+      "loss": 3.5457,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094686519523862e-05,
+      "loss": 3.5249,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09384792477281e-05,
+      "loss": 3.5274,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093009330021758e-05,
+      "loss": 3.526,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092170735270706e-05,
+      "loss": 3.5255,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091333778400027e-05,
+      "loss": 3.5268,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090495183648975e-05,
+      "loss": 3.5444,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089656588897923e-05,
+      "loss": 3.5337,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088817994146871e-05,
+      "loss": 3.5286,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087979399395819e-05,
+      "loss": 3.531,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0871424425251397e-05,
+      "loss": 3.5224,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0863038477740877e-05,
+      "loss": 3.5356,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085465253023036e-05,
+      "loss": 3.5053,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084626658271984e-05,
+      "loss": 3.5295,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083789701401305e-05,
+      "loss": 3.5329,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082951106650253e-05,
+      "loss": 3.5362,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082112511899201e-05,
+      "loss": 3.5263,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081273917148149e-05,
+      "loss": 3.5176,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08043696027747e-05,
+      "loss": 3.5267,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079598365526418e-05,
+      "loss": 3.5146,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078759770775366e-05,
+      "loss": 3.5458,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077921176024314e-05,
+      "loss": 3.5265,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077084219153635e-05,
+      "loss": 3.543,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076245624402583e-05,
+      "loss": 3.5261,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075407029651532e-05,
+      "loss": 3.5165,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07456843490048e-05,
+      "loss": 3.5361,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0737314780298006e-05,
+      "loss": 3.5226,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0728928832787486e-05,
+      "loss": 3.5227,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0720542885276966e-05,
+      "loss": 3.5204,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0712156937766446e-05,
+      "loss": 3.5334,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0703770990255926e-05,
+      "loss": 3.5137,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0695401421549135e-05,
+      "loss": 3.5234,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0687015474038615e-05,
+      "loss": 3.5164,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0678629526528095e-05,
+      "loss": 3.517,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0670243579017575e-05,
+      "loss": 3.5332,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0661874010310784e-05,
+      "loss": 3.5325,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065348806280027e-05,
+      "loss": 3.5234,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064510211528975e-05,
+      "loss": 3.5228,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063671616777923e-05,
+      "loss": 3.5349,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062834659907244e-05,
+      "loss": 3.5269,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061996065156192e-05,
+      "loss": 3.532,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06115747040514e-05,
+      "loss": 3.5269,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060318875654088e-05,
+      "loss": 3.5245,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059481918783409e-05,
+      "loss": 3.5316,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058643324032357e-05,
+      "loss": 3.5274,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057804729281305e-05,
+      "loss": 3.5277,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056966134530253e-05,
+      "loss": 3.536,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056129177659574e-05,
+      "loss": 3.5245,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0552905829085224e-05,
+      "loss": 3.5089,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0544519881574704e-05,
+      "loss": 3.5266,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053613393406418e-05,
+      "loss": 3.5283,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052776436535739e-05,
+      "loss": 3.5256,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051937841784687e-05,
+      "loss": 3.5256,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0510992470336353e-05,
+      "loss": 3.5191,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0502606522825826e-05,
+      "loss": 3.5256,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0494220575315306e-05,
+      "loss": 3.5226,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0485851006608522e-05,
+      "loss": 3.5253,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0477465059098002e-05,
+      "loss": 3.5301,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046907911158748e-05,
+      "loss": 3.5238,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0460709542880695e-05,
+      "loss": 3.5362,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0452323595370175e-05,
+      "loss": 3.5101,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044393764785965e-05,
+      "loss": 3.52,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043555170034913e-05,
+      "loss": 3.5259,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042716575283861e-05,
+      "loss": 3.5268,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0418796184131827e-05,
+      "loss": 3.5288,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04104102366213e-05,
+      "loss": 3.5302,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040202428911078e-05,
+      "loss": 3.5152,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.039363834160026e-05,
+      "loss": 3.5264,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0385268772893476e-05,
+      "loss": 3.5244,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0376882825382952e-05,
+      "loss": 3.5166,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0368496877872432e-05,
+      "loss": 3.5164,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0360110930361912e-05,
+      "loss": 3.5245,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035174136165512e-05,
+      "loss": 3.5275,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03433554141446e-05,
+      "loss": 3.5105,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0334969466634085e-05,
+      "loss": 3.5207,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0326583519123565e-05,
+      "loss": 3.5083,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0318213950416774e-05,
+      "loss": 3.5252,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0309828002906254e-05,
+      "loss": 3.5183,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0301442055395734e-05,
+      "loss": 3.5195,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0293056107885214e-05,
+      "loss": 3.5253,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0284686539178426e-05,
+      "loss": 3.5253,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0276300591667906e-05,
+      "loss": 3.5188,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0267914644157386e-05,
+      "loss": 3.5253,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0259528696646866e-05,
+      "loss": 3.528,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0251142749136346e-05,
+      "loss": 3.5156,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0242773180429555e-05,
+      "loss": 3.5111,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.023438723291904e-05,
+      "loss": 3.5226,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.022600128540852e-05,
+      "loss": 3.5193,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0217615337897998e-05,
+      "loss": 3.5282,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0209245769191207e-05,
+      "loss": 3.5172,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0200859821680687e-05,
+      "loss": 3.5203,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0192473874170167e-05,
+      "loss": 3.5208,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0184087926659647e-05,
+      "loss": 3.517,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.017571835795286e-05,
+      "loss": 3.5105,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.016733241044234e-05,
+      "loss": 3.5287,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015894646293182e-05,
+      "loss": 3.5351,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01505605154213e-05,
+      "loss": 3.5196,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014219094671451e-05,
+      "loss": 3.5118,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0133804999203992e-05,
+      "loss": 3.5266,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0125419051693472e-05,
+      "loss": 3.5232,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0117033104182952e-05,
+      "loss": 3.5154,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010866353547616e-05,
+      "loss": 3.5213,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010027758796564e-05,
+      "loss": 3.5302,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009189164045512e-05,
+      "loss": 3.5247,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.00835056929446e-05,
+      "loss": 3.5246,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0075136124237813e-05,
+      "loss": 3.5152,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0066750176727293e-05,
+      "loss": 3.5239,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0058364229216773e-05,
+      "loss": 3.5224,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0049978281706253e-05,
+      "loss": 3.5122,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0041608712999462e-05,
+      "loss": 3.5412,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0033222765488946e-05,
+      "loss": 3.5174,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0024836817978426e-05,
+      "loss": 3.5257,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0016450870467906e-05,
+      "loss": 3.5222,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0008081301761115e-05,
+      "loss": 3.5208,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.898534059524536,
+      "eval_runtime": 301.7681,
+      "eval_samples_per_second": 1264.517,
+      "eval_steps_per_second": 39.517,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9999695354250595e-05,
+      "loss": 3.5144,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9991309406740075e-05,
+      "loss": 3.517,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9982923459229555e-05,
+      "loss": 3.5309,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9974537511719038e-05,
+      "loss": 3.5223,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9966167943012247e-05,
+      "loss": 3.5247,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9957781995501727e-05,
+      "loss": 3.5137,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9949396047991207e-05,
+      "loss": 3.519,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9941010100480687e-05,
+      "loss": 3.5158,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.99326405317739e-05,
+      "loss": 3.5254,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.992425458426338e-05,
+      "loss": 3.5133,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.991586863675286e-05,
+      "loss": 3.5271,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.990749906804607e-05,
+      "loss": 3.5267,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989911312053555e-05,
+      "loss": 3.5118,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9890727173025028e-05,
+      "loss": 3.5137,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9882341225514508e-05,
+      "loss": 3.502,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987395527800399e-05,
+      "loss": 3.5092,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.986556933049347e-05,
+      "loss": 3.5172,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.985718338298295e-05,
+      "loss": 3.5093,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984879743547243e-05,
+      "loss": 3.5175,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984042786676564e-05,
+      "loss": 3.5357,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983204191925512e-05,
+      "loss": 3.519,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98236559717446e-05,
+      "loss": 3.5193,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.981527002423408e-05,
+      "loss": 3.5181,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9806900455527293e-05,
+      "loss": 3.5148,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9798514508016773e-05,
+      "loss": 3.52,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9790128560506253e-05,
+      "loss": 3.5071,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978174261299573e-05,
+      "loss": 3.5163,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9773373044288945e-05,
+      "loss": 3.5156,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9764987096778425e-05,
+      "loss": 3.5018,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9756601149267905e-05,
+      "loss": 3.5163,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974821520175738e-05,
+      "loss": 3.5117,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9739845633050594e-05,
+      "loss": 3.5203,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9731459685540074e-05,
+      "loss": 3.5175,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9723073738029554e-05,
+      "loss": 3.5089,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.971468779051903e-05,
+      "loss": 3.5151,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9706318221812247e-05,
+      "loss": 3.5322,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9697932274301727e-05,
+      "loss": 3.5048,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96895463267912e-05,
+      "loss": 3.5154,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9681160379280683e-05,
+      "loss": 3.5175,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9672790810573896e-05,
+      "loss": 3.5059,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.966440486306338e-05,
+      "loss": 3.5134,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9656018915552852e-05,
+      "loss": 3.5309,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9647632968042332e-05,
+      "loss": 3.5224,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9639247020531812e-05,
+      "loss": 3.5104,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9630877451825028e-05,
+      "loss": 3.52,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9622491504314504e-05,
+      "loss": 3.5117,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9614105556803984e-05,
+      "loss": 3.5185,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9605719609293464e-05,
+      "loss": 3.4947,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9597350040586673e-05,
+      "loss": 3.5122,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9588964093076153e-05,
+      "loss": 3.5218,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9580578145565633e-05,
+      "loss": 3.5182,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9572192198055117e-05,
+      "loss": 3.5149,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9563822629348326e-05,
+      "loss": 3.5033,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9555436681837806e-05,
+      "loss": 3.5178,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9547050734327286e-05,
+      "loss": 3.5002,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9538664786816766e-05,
+      "loss": 3.529,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9530295218109978e-05,
+      "loss": 3.5122,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9521909270599458e-05,
+      "loss": 3.527,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9513523323088938e-05,
+      "loss": 3.5115,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9505137375578418e-05,
+      "loss": 3.5048,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9496767806871627e-05,
+      "loss": 3.5176,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9488381859361107e-05,
+      "loss": 3.5114,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9479995911850587e-05,
+      "loss": 3.5098,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947160996434007e-05,
+      "loss": 3.5064,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.946324039563328e-05,
+      "loss": 3.5195,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945485444812276e-05,
+      "loss": 3.5015,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.944646850061224e-05,
+      "loss": 3.5064,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943808255310172e-05,
+      "loss": 3.501,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9429712984394932e-05,
+      "loss": 3.5046,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9421327036884412e-05,
+      "loss": 3.5213,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9412941089373892e-05,
+      "loss": 3.5189,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940455514186337e-05,
+      "loss": 3.5089,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.939618557315658e-05,
+      "loss": 3.5088,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.938779962564606e-05,
+      "loss": 3.5172,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937941367813554e-05,
+      "loss": 3.5141,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9371027730625024e-05,
+      "loss": 3.5144,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9362658161918233e-05,
+      "loss": 3.5147,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9354272214407713e-05,
+      "loss": 3.5097,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9345886266897193e-05,
+      "loss": 3.5158,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9337500319386673e-05,
+      "loss": 3.5146,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9329130750679885e-05,
+      "loss": 3.5162,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9320744803169365e-05,
+      "loss": 3.5203,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9312358855658845e-05,
+      "loss": 3.51,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9303972908148325e-05,
+      "loss": 3.4953,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9295603339441534e-05,
+      "loss": 3.5111,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9287217391931014e-05,
+      "loss": 3.514,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9278831444420494e-05,
+      "loss": 3.514,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9270445496909978e-05,
+      "loss": 3.5102,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9262075928203187e-05,
+      "loss": 3.5024,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9253689980692667e-05,
+      "loss": 3.5143,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9245304033182147e-05,
+      "loss": 3.504,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9236918085671627e-05,
+      "loss": 3.5127,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.922854851696484e-05,
+      "loss": 3.5122,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.922016256945432e-05,
+      "loss": 3.5084,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.92117766219438e-05,
+      "loss": 3.5237,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920339067443328e-05,
+      "loss": 3.4995,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9195021105726488e-05,
+      "loss": 3.4989,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9186635158215968e-05,
+      "loss": 3.5129,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9178249210705448e-05,
+      "loss": 3.5133,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916986326319493e-05,
+      "loss": 3.5163,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916149369448814e-05,
+      "loss": 3.5119,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915310774697762e-05,
+      "loss": 3.5014,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91447217994671e-05,
+      "loss": 3.5093,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.913633585195658e-05,
+      "loss": 3.517,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9127966283249793e-05,
+      "loss": 3.5001,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9119580335739273e-05,
+      "loss": 3.501,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9111194388228753e-05,
+      "loss": 3.5144,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9102808440718233e-05,
+      "loss": 3.5138,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9094438872011442e-05,
+      "loss": 3.4965,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9086052924500922e-05,
+      "loss": 3.5011,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90776669769904e-05,
+      "loss": 3.4999,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9069281029479885e-05,
+      "loss": 3.5112,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9060911460773094e-05,
+      "loss": 3.5055,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9052525513262574e-05,
+      "loss": 3.507,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9044139565752054e-05,
+      "loss": 3.5098,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9035753618241534e-05,
+      "loss": 3.5147,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9027384049534746e-05,
+      "loss": 3.5006,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9018998102024226e-05,
+      "loss": 3.513,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9010612154513706e-05,
+      "loss": 3.5123,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9002226207003186e-05,
+      "loss": 3.5049,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8993856638296395e-05,
+      "loss": 3.4931,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8985470690785875e-05,
+      "loss": 3.5089,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8977084743275355e-05,
+      "loss": 3.5062,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896869879576484e-05,
+      "loss": 3.5129,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896031284825432e-05,
+      "loss": 3.5013,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8951943279547528e-05,
+      "loss": 3.5111,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8943557332037008e-05,
+      "loss": 3.5047,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8935171384526488e-05,
+      "loss": 3.5022,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8926785437015968e-05,
+      "loss": 3.4917,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.891841586830918e-05,
+      "loss": 3.5192,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.891002992079866e-05,
+      "loss": 3.5194,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.890164397328814e-05,
+      "loss": 3.5054,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889325802577762e-05,
+      "loss": 3.496,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888488845707083e-05,
+      "loss": 3.5196,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887650250956031e-05,
+      "loss": 3.5066,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886811656204979e-05,
+      "loss": 3.5048,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8859730614539272e-05,
+      "loss": 3.5044,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885136104583248e-05,
+      "loss": 3.5174,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.884297509832196e-05,
+      "loss": 3.5069,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883458915081144e-05,
+      "loss": 3.5146,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882620320330092e-05,
+      "loss": 3.4992,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8817833634594134e-05,
+      "loss": 3.5111,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8809447687083614e-05,
+      "loss": 3.509,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8801061739573094e-05,
+      "loss": 3.4997,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8792675792062574e-05,
+      "loss": 3.5253,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8784306223355783e-05,
+      "loss": 3.502,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8775920275845263e-05,
+      "loss": 3.5105,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8767534328334743e-05,
+      "loss": 3.509,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8759148380824226e-05,
+      "loss": 3.5069,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9002819061279297,
+      "eval_runtime": 305.8485,
+      "eval_samples_per_second": 1247.647,
+      "eval_steps_per_second": 38.99,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8750778812117435e-05,
+      "loss": 3.4963,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8742392864606915e-05,
+      "loss": 3.5064,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8734006917096395e-05,
+      "loss": 3.5141,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8725637348389604e-05,
+      "loss": 3.5091,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8717251400879087e-05,
+      "loss": 3.5152,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8708865453368567e-05,
+      "loss": 3.4987,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8700479505858047e-05,
+      "loss": 3.503,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869209355834752e-05,
+      "loss": 3.5014,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8683723989640736e-05,
+      "loss": 3.513,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8675338042130216e-05,
+      "loss": 3.5016,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8666952094619696e-05,
+      "loss": 3.51,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865858252591291e-05,
+      "loss": 3.5136,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865019657840239e-05,
+      "loss": 3.4968,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.864181063089187e-05,
+      "loss": 3.5009,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.863342468338135e-05,
+      "loss": 3.4894,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8625038735870825e-05,
+      "loss": 3.4991,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8616652788360305e-05,
+      "loss": 3.4979,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8608266840849785e-05,
+      "loss": 3.4972,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8599880893339265e-05,
+      "loss": 3.5008,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8591511324632474e-05,
+      "loss": 3.5204,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8583125377121954e-05,
+      "loss": 3.5031,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8574739429611434e-05,
+      "loss": 3.502,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8566353482100917e-05,
+      "loss": 3.5082,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8557983913394127e-05,
+      "loss": 3.5033,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8549597965883606e-05,
+      "loss": 3.5005,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8541212018373086e-05,
+      "loss": 3.4963,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8532826070862566e-05,
+      "loss": 3.5025,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.852445650215578e-05,
+      "loss": 3.5004,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.851607055464526e-05,
+      "loss": 3.4873,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.850768460713474e-05,
+      "loss": 3.5045,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.849929865962422e-05,
+      "loss": 3.4942,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8490929090917428e-05,
+      "loss": 3.5087,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8482543143406908e-05,
+      "loss": 3.5065,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8474157195896388e-05,
+      "loss": 3.4908,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.846577124838587e-05,
+      "loss": 3.5067,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.845740167967908e-05,
+      "loss": 3.513,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.844901573216856e-05,
+      "loss": 3.4915,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.844062978465804e-05,
+      "loss": 3.5066,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843224383714752e-05,
+      "loss": 3.4988,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8423874268440733e-05,
+      "loss": 3.4935,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8415488320930213e-05,
+      "loss": 3.5017,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8407102373419692e-05,
+      "loss": 3.5178,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8398716425909172e-05,
+      "loss": 3.5049,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.839034685720238e-05,
+      "loss": 3.4987,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838196090969186e-05,
+      "loss": 3.5069,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.837357496218134e-05,
+      "loss": 3.4956,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8365189014670825e-05,
+      "loss": 3.5046,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8356819445964034e-05,
+      "loss": 3.4857,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8348433498453514e-05,
+      "loss": 3.4947,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8340047550942994e-05,
+      "loss": 3.508,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8331661603432474e-05,
+      "loss": 3.5073,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8323292034725686e-05,
+      "loss": 3.4964,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8314906087215166e-05,
+      "loss": 3.488,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8306520139704646e-05,
+      "loss": 3.5072,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8298134192194126e-05,
+      "loss": 3.4835,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8289764623487335e-05,
+      "loss": 3.5122,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8281378675976815e-05,
+      "loss": 3.5016,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8272992728466295e-05,
+      "loss": 3.512,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.826460678095578e-05,
+      "loss": 3.498,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8256237212248988e-05,
+      "loss": 3.4923,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8247851264738467e-05,
+      "loss": 3.4995,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8239465317227947e-05,
+      "loss": 3.5025,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8231079369717427e-05,
+      "loss": 3.4915,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.822270980101064e-05,
+      "loss": 3.4921,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.821432385350012e-05,
+      "loss": 3.5064,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.82059379059896e-05,
+      "loss": 3.4933,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.819755195847908e-05,
+      "loss": 3.4868,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818918238977229e-05,
+      "loss": 3.4887,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818079644226177e-05,
+      "loss": 3.4914,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817241049475125e-05,
+      "loss": 3.5057,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8164024547240732e-05,
+      "loss": 3.5063,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.815565497853394e-05,
+      "loss": 3.4939,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.814726903102342e-05,
+      "loss": 3.4965,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81388830835129e-05,
+      "loss": 3.502,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.813049713600238e-05,
+      "loss": 3.5006,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8122127567295594e-05,
+      "loss": 3.5,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8113741619785074e-05,
+      "loss": 3.5026,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8105355672274553e-05,
+      "loss": 3.4976,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8096969724764033e-05,
+      "loss": 3.5039,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8088600156057243e-05,
+      "loss": 3.4991,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8080214208546722e-05,
+      "loss": 3.5006,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8071828261036202e-05,
+      "loss": 3.5037,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8063442313525686e-05,
+      "loss": 3.5001,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8055072744818895e-05,
+      "loss": 3.4846,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8046686797308375e-05,
+      "loss": 3.4973,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8038300849797855e-05,
+      "loss": 3.4987,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8029914902287335e-05,
+      "loss": 3.5006,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8021545333580547e-05,
+      "loss": 3.4991,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8013159386070027e-05,
+      "loss": 3.4867,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8004773438559507e-05,
+      "loss": 3.5008,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7996387491048987e-05,
+      "loss": 3.4919,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7988017922342196e-05,
+      "loss": 3.5014,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7979631974831676e-05,
+      "loss": 3.4926,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7971246027321156e-05,
+      "loss": 3.4973,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7962860079810636e-05,
+      "loss": 3.5037,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795449051110385e-05,
+      "loss": 3.4984,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.794610456359333e-05,
+      "loss": 3.4785,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.793771861608281e-05,
+      "loss": 3.4997,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.792933266857229e-05,
+      "loss": 3.4995,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7920963099865497e-05,
+      "loss": 3.5012,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.791257715235498e-05,
+      "loss": 3.5006,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.790419120484446e-05,
+      "loss": 3.4907,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.789580525733394e-05,
+      "loss": 3.4908,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.788743568862715e-05,
+      "loss": 3.5039,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.787904974111663e-05,
+      "loss": 3.4838,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.787066379360611e-05,
+      "loss": 3.4915,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786227784609559e-05,
+      "loss": 3.5007,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7853908277388802e-05,
+      "loss": 3.5042,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7845522329878282e-05,
+      "loss": 3.4784,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7837136382367762e-05,
+      "loss": 3.4954,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7828750434857242e-05,
+      "loss": 3.4813,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.782038086615045e-05,
+      "loss": 3.5002,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7811994918639935e-05,
+      "loss": 3.4922,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7803608971129414e-05,
+      "loss": 3.4859,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7795223023618894e-05,
+      "loss": 3.5042,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7786853454912104e-05,
+      "loss": 3.4982,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7778467507401583e-05,
+      "loss": 3.4886,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7770081559891063e-05,
+      "loss": 3.5007,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7761695612380543e-05,
+      "loss": 3.4976,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7753326043673756e-05,
+      "loss": 3.4918,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7744940096163236e-05,
+      "loss": 3.4784,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7736554148652716e-05,
+      "loss": 3.5013,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7728168201142196e-05,
+      "loss": 3.4903,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7719798632435405e-05,
+      "loss": 3.4974,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7711412684924888e-05,
+      "loss": 3.489,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7703026737414368e-05,
+      "loss": 3.4965,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7694640789903848e-05,
+      "loss": 3.4926,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7686271221197057e-05,
+      "loss": 3.4906,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7677885273686537e-05,
+      "loss": 3.477,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7669499326176017e-05,
+      "loss": 3.509,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7661113378665494e-05,
+      "loss": 3.5002,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765274380995871e-05,
+      "loss": 3.4941,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.764435786244819e-05,
+      "loss": 3.4845,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.763597191493767e-05,
+      "loss": 3.4995,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7627585967427143e-05,
+      "loss": 3.4983,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.761921639872036e-05,
+      "loss": 3.4878,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7610830451209842e-05,
+      "loss": 3.4885,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7602444503699322e-05,
+      "loss": 3.5007,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7594058556188795e-05,
+      "loss": 3.4929,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758568898748201e-05,
+      "loss": 3.5061,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757730303997149e-05,
+      "loss": 3.4863,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7568917092460967e-05,
+      "loss": 3.4972,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7560531144950447e-05,
+      "loss": 3.4953,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7552161576243663e-05,
+      "loss": 3.486,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7543775628733143e-05,
+      "loss": 3.5094,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7535389681222616e-05,
+      "loss": 3.4922,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7527003733712096e-05,
+      "loss": 3.4968,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7518634165005312e-05,
+      "loss": 3.4892,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7510248217494796e-05,
+      "loss": 3.4961,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.902003526687622,
+      "eval_runtime": 305.6413,
+      "eval_samples_per_second": 1248.493,
+      "eval_steps_per_second": 39.016,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.750186226998427e-05,
+      "loss": 3.4824,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.749347632247375e-05,
+      "loss": 3.4968,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.748509037496323e-05,
+      "loss": 3.4994,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.747670442745271e-05,
+      "loss": 3.4947,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.746833485874592e-05,
+      "loss": 3.5046,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.74599489112354e-05,
+      "loss": 3.4801,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.745156296372488e-05,
+      "loss": 3.4888,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744317701621436e-05,
+      "loss": 3.4884,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743480744750757e-05,
+      "loss": 3.4993,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742642149999705e-05,
+      "loss": 3.4901,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7418035552486533e-05,
+      "loss": 3.4965,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7409649604976013e-05,
+      "loss": 3.4965,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7401280036269222e-05,
+      "loss": 3.4844,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7392894088758702e-05,
+      "loss": 3.4882,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7384508141248182e-05,
+      "loss": 3.4743,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7376122193737662e-05,
+      "loss": 3.4841,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7367736246227142e-05,
+      "loss": 3.4854,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7359350298716626e-05,
+      "loss": 3.4859,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7350964351206105e-05,
+      "loss": 3.4862,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7342594782499315e-05,
+      "loss": 3.5089,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7334208834988795e-05,
+      "loss": 3.4921,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7325822887478274e-05,
+      "loss": 3.49,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7317436939967754e-05,
+      "loss": 3.4951,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7309067371260967e-05,
+      "loss": 3.4928,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7300681423750447e-05,
+      "loss": 3.4834,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7292295476239927e-05,
+      "loss": 3.4836,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7283909528729407e-05,
+      "loss": 3.4933,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7275523581218887e-05,
+      "loss": 3.4822,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7267154012512096e-05,
+      "loss": 3.4764,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.725876806500158e-05,
+      "loss": 3.4909,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.725038211749106e-05,
+      "loss": 3.483,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724199616998054e-05,
+      "loss": 3.4951,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7233626601273748e-05,
+      "loss": 3.4926,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7225240653763228e-05,
+      "loss": 3.4807,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7216854706252708e-05,
+      "loss": 3.4905,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720848513754592e-05,
+      "loss": 3.5036,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72000991900354e-05,
+      "loss": 3.4765,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719171324252488e-05,
+      "loss": 3.4927,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718332729501436e-05,
+      "loss": 3.4818,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.717495772630757e-05,
+      "loss": 3.4851,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.716657177879705e-05,
+      "loss": 3.4846,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.715818583128653e-05,
+      "loss": 3.505,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7149799883776013e-05,
+      "loss": 3.4918,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7141430315069222e-05,
+      "loss": 3.4869,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7133044367558702e-05,
+      "loss": 3.4925,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7124658420048182e-05,
+      "loss": 3.4815,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7116272472537662e-05,
+      "loss": 3.4919,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7107902903830874e-05,
+      "loss": 3.4734,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7099516956320354e-05,
+      "loss": 3.4791,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7091131008809834e-05,
+      "loss": 3.4929,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7082745061299314e-05,
+      "loss": 3.4903,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7074375492592523e-05,
+      "loss": 3.4838,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7065989545082003e-05,
+      "loss": 3.4795,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7057603597571483e-05,
+      "loss": 3.4918,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7049217650060966e-05,
+      "loss": 3.4717,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7040848081354176e-05,
+      "loss": 3.4961,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7032462133843656e-05,
+      "loss": 3.4915,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7024076186333135e-05,
+      "loss": 3.4981,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7015690238822615e-05,
+      "loss": 3.4918,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7007320670115828e-05,
+      "loss": 3.4794,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6998934722605308e-05,
+      "loss": 3.4845,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6990548775094788e-05,
+      "loss": 3.4891,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6982162827584268e-05,
+      "loss": 3.4775,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6973793258877477e-05,
+      "loss": 3.4787,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6965407311366957e-05,
+      "loss": 3.49,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6957021363856437e-05,
+      "loss": 3.4826,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.694863541634592e-05,
+      "loss": 3.4726,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.694026584763913e-05,
+      "loss": 3.4789,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693187990012861e-05,
+      "loss": 3.4771,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.692349395261809e-05,
+      "loss": 3.4918,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.691510800510757e-05,
+      "loss": 3.4921,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.690673843640078e-05,
+      "loss": 3.4828,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689835248889026e-05,
+      "loss": 3.4873,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.688996654137974e-05,
+      "loss": 3.4832,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.688158059386922e-05,
+      "loss": 3.4901,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.687321102516243e-05,
+      "loss": 3.484,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.686482507765191e-05,
+      "loss": 3.4915,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.685643913014139e-05,
+      "loss": 3.4865,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6848053182630874e-05,
+      "loss": 3.4867,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6839683613924083e-05,
+      "loss": 3.4867,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6831297666413563e-05,
+      "loss": 3.4884,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6822911718903043e-05,
+      "loss": 3.4898,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.681452577139252e-05,
+      "loss": 3.4891,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6806156202685735e-05,
+      "loss": 3.4742,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6797770255175215e-05,
+      "loss": 3.4823,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6789384307664695e-05,
+      "loss": 3.4869,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.678099836015417e-05,
+      "loss": 3.4871,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6772628791447384e-05,
+      "loss": 3.4918,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6764242843936864e-05,
+      "loss": 3.4732,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.675585689642634e-05,
+      "loss": 3.4838,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.674747094891582e-05,
+      "loss": 3.482,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6739101380209037e-05,
+      "loss": 3.49,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6730715432698517e-05,
+      "loss": 3.4763,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.672232948518799e-05,
+      "loss": 3.4859,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6713943537677473e-05,
+      "loss": 3.4934,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6705590347774418e-05,
+      "loss": 3.4835,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6697204400263898e-05,
+      "loss": 3.4677,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6688818452753378e-05,
+      "loss": 3.4847,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6680432505242858e-05,
+      "loss": 3.4881,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6672046557732338e-05,
+      "loss": 3.4878,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6663660610221814e-05,
+      "loss": 3.487,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6655274662711294e-05,
+      "loss": 3.4765,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6646888715200774e-05,
+      "loss": 3.4805,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.663851914649399e-05,
+      "loss": 3.4912,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6630133198983463e-05,
+      "loss": 3.4745,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6621747251472943e-05,
+      "loss": 3.4796,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6613361303962427e-05,
+      "loss": 3.4852,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6604991735255643e-05,
+      "loss": 3.4902,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6596605787745116e-05,
+      "loss": 3.4645,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6588219840234596e-05,
+      "loss": 3.4851,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6579833892724076e-05,
+      "loss": 3.4694,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6571464324017288e-05,
+      "loss": 3.4827,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6563078376506768e-05,
+      "loss": 3.4824,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6554692428996248e-05,
+      "loss": 3.475,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6546306481485728e-05,
+      "loss": 3.4894,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6537936912778937e-05,
+      "loss": 3.49,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6529550965268417e-05,
+      "loss": 3.4708,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6521165017757897e-05,
+      "loss": 3.486,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651277907024738e-05,
+      "loss": 3.4866,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.650440950154059e-05,
+      "loss": 3.4811,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.649602355403007e-05,
+      "loss": 3.4674,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648763760651955e-05,
+      "loss": 3.4855,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.647925165900903e-05,
+      "loss": 3.4804,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6470882090302242e-05,
+      "loss": 3.4849,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6462496142791722e-05,
+      "loss": 3.4783,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6454110195281202e-05,
+      "loss": 3.4856,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6445724247770682e-05,
+      "loss": 3.4742,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643735467906389e-05,
+      "loss": 3.4765,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642896873155337e-05,
+      "loss": 3.4674,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642058278404285e-05,
+      "loss": 3.4942,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6412196836532334e-05,
+      "loss": 3.4846,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6403827267825543e-05,
+      "loss": 3.4845,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6395441320315023e-05,
+      "loss": 3.4728,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6387055372804503e-05,
+      "loss": 3.4894,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6378669425293983e-05,
+      "loss": 3.487,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6370283477783463e-05,
+      "loss": 3.4736,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6361913909076675e-05,
+      "loss": 3.4787,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6353527961566155e-05,
+      "loss": 3.487,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6345142014055635e-05,
+      "loss": 3.482,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6336772445348844e-05,
+      "loss": 3.4932,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6328386497838324e-05,
+      "loss": 3.4751,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6320000550327804e-05,
+      "loss": 3.4832,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6311614602817288e-05,
+      "loss": 3.478,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6303245034110497e-05,
+      "loss": 3.4743,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6294859086599977e-05,
+      "loss": 3.4977,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6286473139089457e-05,
+      "loss": 3.4762,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6278087191578937e-05,
+      "loss": 3.4844,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.626971762287215e-05,
+      "loss": 3.4794,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.626133167536163e-05,
+      "loss": 3.4837,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9046437740325928,
+      "eval_runtime": 303.0693,
+      "eval_samples_per_second": 1259.088,
+      "eval_steps_per_second": 39.347,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.625294572785111e-05,
+      "loss": 3.4774,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.624455978034059e-05,
+      "loss": 3.4827,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6236190211633798e-05,
+      "loss": 3.4845,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6227804264123278e-05,
+      "loss": 3.483,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6219418316612758e-05,
+      "loss": 3.4917,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.621103236910224e-05,
+      "loss": 3.4737,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.620266280039545e-05,
+      "loss": 3.476,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.619427685288493e-05,
+      "loss": 3.4696,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.618589090537441e-05,
+      "loss": 3.49,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.617750495786389e-05,
+      "loss": 3.4742,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6169135389157103e-05,
+      "loss": 3.4844,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6160749441646583e-05,
+      "loss": 3.4859,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6152379872939792e-05,
+      "loss": 3.477,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6143993925429272e-05,
+      "loss": 3.4679,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6135607977918752e-05,
+      "loss": 3.467,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6127222030408232e-05,
+      "loss": 3.4676,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6118836082897712e-05,
+      "loss": 3.4742,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.611045013538719e-05,
+      "loss": 3.4753,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6102064187876675e-05,
+      "loss": 3.4708,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6093678240366155e-05,
+      "loss": 3.4933,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6085308671659364e-05,
+      "loss": 3.4802,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6076922724148844e-05,
+      "loss": 3.4771,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6068536776638324e-05,
+      "loss": 3.4838,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6060150829127804e-05,
+      "loss": 3.4814,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6051781260421016e-05,
+      "loss": 3.4728,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6043395312910496e-05,
+      "loss": 3.4698,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6035009365399976e-05,
+      "loss": 3.485,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6026639796693185e-05,
+      "loss": 3.4653,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6018253849182665e-05,
+      "loss": 3.468,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6009867901672145e-05,
+      "loss": 3.4723,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600148195416163e-05,
+      "loss": 3.4705,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5993112385454838e-05,
+      "loss": 3.4846,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5984726437944318e-05,
+      "loss": 3.4758,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5976340490433798e-05,
+      "loss": 3.4687,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5967954542923278e-05,
+      "loss": 3.4795,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.595958497421649e-05,
+      "loss": 3.4884,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.595119902670597e-05,
+      "loss": 3.4676,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.594281307919545e-05,
+      "loss": 3.4774,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.593442713168493e-05,
+      "loss": 3.4658,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.592605756297814e-05,
+      "loss": 3.4795,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.591767161546762e-05,
+      "loss": 3.4668,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.59092856679571e-05,
+      "loss": 3.4936,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5900899720446582e-05,
+      "loss": 3.4761,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5892513772936062e-05,
+      "loss": 3.4747,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.588414420422927e-05,
+      "loss": 3.4807,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.587575825671875e-05,
+      "loss": 3.4706,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.586737230920823e-05,
+      "loss": 3.4793,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5859002740501444e-05,
+      "loss": 3.4637,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5850616792990924e-05,
+      "loss": 3.4643,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5842230845480404e-05,
+      "loss": 3.4823,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5833844897969884e-05,
+      "loss": 3.4751,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5825475329263093e-05,
+      "loss": 3.4774,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5817089381752573e-05,
+      "loss": 3.4665,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5808703434242053e-05,
+      "loss": 3.4759,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5800317486731536e-05,
+      "loss": 3.4592,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5791947918024745e-05,
+      "loss": 3.4821,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5783561970514225e-05,
+      "loss": 3.4753,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5775176023003705e-05,
+      "loss": 3.4866,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5766790075493185e-05,
+      "loss": 3.4775,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5758420506786397e-05,
+      "loss": 3.4695,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5750034559275877e-05,
+      "loss": 3.4686,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5741648611765357e-05,
+      "loss": 3.4787,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5733262664254837e-05,
+      "loss": 3.4681,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5724893095548046e-05,
+      "loss": 3.4674,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5716507148037526e-05,
+      "loss": 3.4734,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5708121200527006e-05,
+      "loss": 3.4745,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569973525301649e-05,
+      "loss": 3.4578,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.56913656843097e-05,
+      "loss": 3.4646,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568297973679918e-05,
+      "loss": 3.4638,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.567459378928866e-05,
+      "loss": 3.4817,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.566620784177814e-05,
+      "loss": 3.4766,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.565783827307135e-05,
+      "loss": 3.4686,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.564945232556083e-05,
+      "loss": 3.477,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.564106637805031e-05,
+      "loss": 3.4715,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5632680430539784e-05,
+      "loss": 3.4848,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5624310861833e-05,
+      "loss": 3.4655,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.561592491432248e-05,
+      "loss": 3.4849,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.560753896681196e-05,
+      "loss": 3.469,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5599153019301437e-05,
+      "loss": 3.4762,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5590783450594652e-05,
+      "loss": 3.4797,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5582397503084132e-05,
+      "loss": 3.4727,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5574011555573612e-05,
+      "loss": 3.4802,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556562560806309e-05,
+      "loss": 3.4733,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5557256039356305e-05,
+      "loss": 3.4627,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5548870091845785e-05,
+      "loss": 3.4726,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5540484144335258e-05,
+      "loss": 3.4676,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5532098196824738e-05,
+      "loss": 3.4781,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5523728628117954e-05,
+      "loss": 3.4753,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5515342680607434e-05,
+      "loss": 3.4615,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.550695673309691e-05,
+      "loss": 3.4763,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.549857078558639e-05,
+      "loss": 3.4666,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5490201216879606e-05,
+      "loss": 3.4765,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5481815269369086e-05,
+      "loss": 3.464,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.547342932185856e-05,
+      "loss": 3.4751,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5465043374348043e-05,
+      "loss": 3.4814,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.545667380564126e-05,
+      "loss": 3.4709,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5448304236934468e-05,
+      "loss": 3.4577,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5439918289423948e-05,
+      "loss": 3.4751,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5431532341913427e-05,
+      "loss": 3.4665,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5423146394402907e-05,
+      "loss": 3.4804,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5414760446892384e-05,
+      "loss": 3.4743,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5406374499381864e-05,
+      "loss": 3.4672,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5397988551871344e-05,
+      "loss": 3.4679,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5389618983164553e-05,
+      "loss": 3.4791,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5381233035654033e-05,
+      "loss": 3.4571,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5372847088143513e-05,
+      "loss": 3.4694,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5364461140632996e-05,
+      "loss": 3.4694,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5356091571926205e-05,
+      "loss": 3.4794,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5347705624415685e-05,
+      "loss": 3.4498,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5339319676905165e-05,
+      "loss": 3.4722,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5330933729394645e-05,
+      "loss": 3.4629,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5322564160687858e-05,
+      "loss": 3.4649,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5314178213177338e-05,
+      "loss": 3.4741,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5305792265666818e-05,
+      "loss": 3.4613,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5297406318156298e-05,
+      "loss": 3.4794,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5289036749449507e-05,
+      "loss": 3.4736,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5280650801938987e-05,
+      "loss": 3.4613,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5272264854428467e-05,
+      "loss": 3.4732,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.526387890691795e-05,
+      "loss": 3.4728,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.525550933821116e-05,
+      "loss": 3.4705,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.524712339070064e-05,
+      "loss": 3.4508,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.523873744319012e-05,
+      "loss": 3.4773,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.52303514956796e-05,
+      "loss": 3.4653,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.522198192697281e-05,
+      "loss": 3.4696,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.521359597946229e-05,
+      "loss": 3.4681,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.520521003195177e-05,
+      "loss": 3.4761,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.519682408444125e-05,
+      "loss": 3.4665,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518845451573446e-05,
+      "loss": 3.4616,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518006856822394e-05,
+      "loss": 3.457,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.517168262071342e-05,
+      "loss": 3.4778,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.51632966732029e-05,
+      "loss": 3.4727,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5154927104496113e-05,
+      "loss": 3.4719,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5146541156985593e-05,
+      "loss": 3.4615,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5138155209475073e-05,
+      "loss": 3.474,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5129769261964553e-05,
+      "loss": 3.4726,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.512139969325776e-05,
+      "loss": 3.4639,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5113013745747245e-05,
+      "loss": 3.4644,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5104627798236725e-05,
+      "loss": 3.4727,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5096241850726205e-05,
+      "loss": 3.4726,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5087872282019414e-05,
+      "loss": 3.4805,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5079486334508894e-05,
+      "loss": 3.4601,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5071100386998374e-05,
+      "loss": 3.4725,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5062714439487854e-05,
+      "loss": 3.4654,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5054344870781066e-05,
+      "loss": 3.4657,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5045958923270546e-05,
+      "loss": 3.4778,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5037572975760026e-05,
+      "loss": 3.4674,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5029187028249506e-05,
+      "loss": 3.4733,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5020817459542715e-05,
+      "loss": 3.4638,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.50124315120322e-05,
+      "loss": 3.4734,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9056832790374756,
+      "eval_runtime": 303.0368,
+      "eval_samples_per_second": 1259.223,
+      "eval_steps_per_second": 39.352,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.500404556452168e-05,
+      "loss": 3.4588,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.499565961701116e-05,
+      "loss": 3.4724,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.498727366950064e-05,
+      "loss": 3.4712,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.497888772199012e-05,
+      "loss": 3.4685,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4970518153283328e-05,
+      "loss": 3.4818,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4962132205772808e-05,
+      "loss": 3.4599,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.495374625826229e-05,
+      "loss": 3.466,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.494536031075177e-05,
+      "loss": 3.4588,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.493699074204498e-05,
+      "loss": 3.4748,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.492860479453446e-05,
+      "loss": 3.4703,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.492021884702394e-05,
+      "loss": 3.4689,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.491183289951342e-05,
+      "loss": 3.4686,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4903479709610365e-05,
+      "loss": 3.4654,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.489509376209984e-05,
+      "loss": 3.4574,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.488670781458932e-05,
+      "loss": 3.4619,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.48783218670788e-05,
+      "loss": 3.4517,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486993591956828e-05,
+      "loss": 3.4648,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486154997205776e-05,
+      "loss": 3.4615,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4853164024547245e-05,
+      "loss": 3.455,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.484477807703672e-05,
+      "loss": 3.4867,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4836408508329934e-05,
+      "loss": 3.4657,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4828022560819414e-05,
+      "loss": 3.4658,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4819636613308894e-05,
+      "loss": 3.4709,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.481125066579837e-05,
+      "loss": 3.4665,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4802881097091586e-05,
+      "loss": 3.4608,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4794495149581066e-05,
+      "loss": 3.4587,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4786109202070542e-05,
+      "loss": 3.4709,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4777723254560022e-05,
+      "loss": 3.449,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4769353685853235e-05,
+      "loss": 3.4588,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4760967738342715e-05,
+      "loss": 3.4574,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4752581790832195e-05,
+      "loss": 3.4571,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4744195843321675e-05,
+      "loss": 3.4694,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4735826274614887e-05,
+      "loss": 3.4671,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4727440327104367e-05,
+      "loss": 3.4606,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4719054379593844e-05,
+      "loss": 3.4681,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.471068481088706e-05,
+      "loss": 3.4725,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.470229886337654e-05,
+      "loss": 3.456,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4693912915866016e-05,
+      "loss": 3.4668,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4685526968355496e-05,
+      "loss": 3.4531,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.467715739964871e-05,
+      "loss": 3.4668,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.466877145213819e-05,
+      "loss": 3.4527,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.466038550462767e-05,
+      "loss": 3.4822,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465199955711715e-05,
+      "loss": 3.4647,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464362998841036e-05,
+      "loss": 3.4612,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.463524404089984e-05,
+      "loss": 3.4692,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4626858093389317e-05,
+      "loss": 3.4592,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4618472145878797e-05,
+      "loss": 3.4646,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4610102577172013e-05,
+      "loss": 3.4533,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.460171662966149e-05,
+      "loss": 3.4499,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459333068215097e-05,
+      "loss": 3.47,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.458494473464045e-05,
+      "loss": 3.4626,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4576575165933662e-05,
+      "loss": 3.4631,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.456818921842314e-05,
+      "loss": 3.4553,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4559803270912622e-05,
+      "loss": 3.4633,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4551417323402102e-05,
+      "loss": 3.4503,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4543047754695315e-05,
+      "loss": 3.4677,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.453466180718479e-05,
+      "loss": 3.464,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.452627585967427e-05,
+      "loss": 3.4757,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.451788991216375e-05,
+      "loss": 3.4646,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4509520343456964e-05,
+      "loss": 3.4555,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4501134395946444e-05,
+      "loss": 3.4599,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4492748448435924e-05,
+      "loss": 3.4626,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4484362500925403e-05,
+      "loss": 3.4537,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4475992932218613e-05,
+      "loss": 3.459,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4467606984708093e-05,
+      "loss": 3.4594,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4459221037197576e-05,
+      "loss": 3.4634,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4450835089687056e-05,
+      "loss": 3.444,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4442465520980265e-05,
+      "loss": 3.4536,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4434079573469745e-05,
+      "loss": 3.4504,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4425693625959225e-05,
+      "loss": 3.4677,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4417307678448705e-05,
+      "loss": 3.4632,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4408938109741917e-05,
+      "loss": 3.4579,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4400552162231397e-05,
+      "loss": 3.4651,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4392166214720877e-05,
+      "loss": 3.4568,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4383780267210357e-05,
+      "loss": 3.4751,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4375410698503566e-05,
+      "loss": 3.4525,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4367024750993046e-05,
+      "loss": 3.4695,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.435863880348253e-05,
+      "loss": 3.4615,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.435026923477574e-05,
+      "loss": 3.4622,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.434188328726522e-05,
+      "loss": 3.4671,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.43334973397547e-05,
+      "loss": 3.4613,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.432511139224418e-05,
+      "loss": 3.4657,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.431672544473366e-05,
+      "loss": 3.4636,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.430833949722314e-05,
+      "loss": 3.4494,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.429995354971262e-05,
+      "loss": 3.4607,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.42915676022021e-05,
+      "loss": 3.4571,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428319803349531e-05,
+      "loss": 3.4633,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.427481208598479e-05,
+      "loss": 3.4614,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4266426138474267e-05,
+      "loss": 3.4501,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4258040190963747e-05,
+      "loss": 3.4616,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4249670622256963e-05,
+      "loss": 3.4578,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4241284674746443e-05,
+      "loss": 3.4643,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423289872723592e-05,
+      "loss": 3.4484,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.42245127797254e-05,
+      "loss": 3.4643,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4216143211018612e-05,
+      "loss": 3.473,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4207757263508092e-05,
+      "loss": 3.4565,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4199387694801305e-05,
+      "loss": 3.4427,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4191001747290785e-05,
+      "loss": 3.4607,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4182615799780264e-05,
+      "loss": 3.4606,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.417422985226974e-05,
+      "loss": 3.4657,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.416584390475922e-05,
+      "loss": 3.4635,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.41574579572487e-05,
+      "loss": 3.4557,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4149072009738184e-05,
+      "loss": 3.4565,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4140702441031393e-05,
+      "loss": 3.4664,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4132316493520873e-05,
+      "loss": 3.4444,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4123930546010353e-05,
+      "loss": 3.4575,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4115544598499833e-05,
+      "loss": 3.4574,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4107175029793046e-05,
+      "loss": 3.4655,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4098789082282526e-05,
+      "loss": 3.4383,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4090403134772006e-05,
+      "loss": 3.4596,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4082017187261486e-05,
+      "loss": 3.4497,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4073647618554695e-05,
+      "loss": 3.4569,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4065261671044175e-05,
+      "loss": 3.4569,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4056875723533655e-05,
+      "loss": 3.4524,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4048489776023138e-05,
+      "loss": 3.4628,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4040120207316347e-05,
+      "loss": 3.4636,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4031734259805827e-05,
+      "loss": 3.4475,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4023348312295307e-05,
+      "loss": 3.4628,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4014962364784787e-05,
+      "loss": 3.4564,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4006592796078e-05,
+      "loss": 3.4608,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.399820684856748e-05,
+      "loss": 3.4414,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.398982090105696e-05,
+      "loss": 3.4638,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.398143495354644e-05,
+      "loss": 3.4483,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397306538483965e-05,
+      "loss": 3.4617,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.396467943732913e-05,
+      "loss": 3.4507,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3956293489818608e-05,
+      "loss": 3.4652,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.394790754230809e-05,
+      "loss": 3.4568,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.39395379736013e-05,
+      "loss": 3.4503,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393115202609078e-05,
+      "loss": 3.4418,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392276607858026e-05,
+      "loss": 3.4696,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.391438013106974e-05,
+      "loss": 3.4561,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3906010562362953e-05,
+      "loss": 3.4616,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3897624614852433e-05,
+      "loss": 3.4504,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3889238667341913e-05,
+      "loss": 3.4626,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3880852719831393e-05,
+      "loss": 3.4643,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3872483151124602e-05,
+      "loss": 3.4496,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3864097203614082e-05,
+      "loss": 3.4511,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3855711256103562e-05,
+      "loss": 3.4605,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3847325308593042e-05,
+      "loss": 3.4615,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3838955739886254e-05,
+      "loss": 3.468,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3830569792375734e-05,
+      "loss": 3.4481,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3822183844865214e-05,
+      "loss": 3.4581,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381379789735469e-05,
+      "loss": 3.4529,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3805428328647907e-05,
+      "loss": 3.4574,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3797042381137387e-05,
+      "loss": 3.4641,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3788656433626867e-05,
+      "loss": 3.4514,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3780270486116343e-05,
+      "loss": 3.4639,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3771900917409556e-05,
+      "loss": 3.4472,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3763514969899036e-05,
+      "loss": 3.467,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.907168388366699,
+      "eval_runtime": 303.0694,
+      "eval_samples_per_second": 1259.088,
+      "eval_steps_per_second": 39.347,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3755129022388516e-05,
+      "loss": 3.4452,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3746743074877996e-05,
+      "loss": 3.463,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3738357127367476e-05,
+      "loss": 3.4562,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3729971179856955e-05,
+      "loss": 3.4575,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3721601611150165e-05,
+      "loss": 3.469,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3713215663639645e-05,
+      "loss": 3.449,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3704829716129124e-05,
+      "loss": 3.4508,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3696443768618608e-05,
+      "loss": 3.4442,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3688074199911817e-05,
+      "loss": 3.4641,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3679688252401297e-05,
+      "loss": 3.4614,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3671302304890777e-05,
+      "loss": 3.4541,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3662916357380257e-05,
+      "loss": 3.4596,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.365454678867347e-05,
+      "loss": 3.4499,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.364616084116295e-05,
+      "loss": 3.4464,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.363777489365243e-05,
+      "loss": 3.4506,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.362938894614191e-05,
+      "loss": 3.438,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3621019377435118e-05,
+      "loss": 3.451,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3612633429924598e-05,
+      "loss": 3.4481,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3604247482414078e-05,
+      "loss": 3.4424,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.359586153490356e-05,
+      "loss": 3.4781,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.358749196619677e-05,
+      "loss": 3.4531,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357910601868625e-05,
+      "loss": 3.4564,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357072007117573e-05,
+      "loss": 3.4591,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356233412366521e-05,
+      "loss": 3.4546,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3553964554958423e-05,
+      "loss": 3.4452,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3545578607447903e-05,
+      "loss": 3.4524,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3537192659937383e-05,
+      "loss": 3.4549,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3528823091230592e-05,
+      "loss": 3.4409,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3520437143720072e-05,
+      "loss": 3.4451,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3512051196209552e-05,
+      "loss": 3.4431,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3503665248699032e-05,
+      "loss": 3.4517,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3495295679992244e-05,
+      "loss": 3.4529,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3486909732481724e-05,
+      "loss": 3.4547,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3478523784971204e-05,
+      "loss": 3.4516,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3470137837460684e-05,
+      "loss": 3.4545,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3461768268753893e-05,
+      "loss": 3.4566,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3453382321243377e-05,
+      "loss": 3.4456,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3444996373732857e-05,
+      "loss": 3.4588,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3436610426222337e-05,
+      "loss": 3.4391,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3428240857515546e-05,
+      "loss": 3.458,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3419854910005026e-05,
+      "loss": 3.437,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3411468962494506e-05,
+      "loss": 3.4746,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3403083014983985e-05,
+      "loss": 3.454,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3394713446277198e-05,
+      "loss": 3.4493,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3386327498766678e-05,
+      "loss": 3.4586,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3377941551256158e-05,
+      "loss": 3.4527,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3369555603745638e-05,
+      "loss": 3.4504,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3361186035038847e-05,
+      "loss": 3.4399,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.335280008752833e-05,
+      "loss": 3.4324,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.334441414001781e-05,
+      "loss": 3.4642,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.333602819250729e-05,
+      "loss": 3.4501,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.33276586238005e-05,
+      "loss": 3.4569,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.331927267628998e-05,
+      "loss": 3.446,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.331088672877946e-05,
+      "loss": 3.4512,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.330250078126894e-05,
+      "loss": 3.4407,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.329413121256215e-05,
+      "loss": 3.4533,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328574526505163e-05,
+      "loss": 3.4502,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.327735931754111e-05,
+      "loss": 3.4641,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3268973370030588e-05,
+      "loss": 3.4569,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.32606038013238e-05,
+      "loss": 3.4422,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3252217853813284e-05,
+      "loss": 3.4475,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3243831906302764e-05,
+      "loss": 3.4554,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.323544595879224e-05,
+      "loss": 3.4398,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3227076390085453e-05,
+      "loss": 3.4456,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3218690442574933e-05,
+      "loss": 3.4503,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321030449506441e-05,
+      "loss": 3.4497,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3201918547553893e-05,
+      "loss": 3.4348,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3193548978847105e-05,
+      "loss": 3.442,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3185163031336585e-05,
+      "loss": 3.439,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3176777083826062e-05,
+      "loss": 3.4547,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3168391136315542e-05,
+      "loss": 3.4519,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3160021567608754e-05,
+      "loss": 3.447,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3151635620098238e-05,
+      "loss": 3.4563,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3143249672587714e-05,
+      "loss": 3.4424,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3134863725077194e-05,
+      "loss": 3.4639,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3126494156370407e-05,
+      "loss": 3.4399,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3118108208859883e-05,
+      "loss": 3.4565,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3109722261349363e-05,
+      "loss": 3.4491,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3101336313838846e-05,
+      "loss": 3.449,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309296674513206e-05,
+      "loss": 3.4574,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3084580797621536e-05,
+      "loss": 3.4488,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3076194850111016e-05,
+      "loss": 3.4539,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3067808902600495e-05,
+      "loss": 3.4533,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3059439333893708e-05,
+      "loss": 3.437,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3051053386383188e-05,
+      "loss": 3.4507,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3042667438872668e-05,
+      "loss": 3.4471,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3034281491362148e-05,
+      "loss": 3.4551,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3025911922655357e-05,
+      "loss": 3.4469,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3017525975144837e-05,
+      "loss": 3.4398,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3009140027634317e-05,
+      "loss": 3.449,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.30007540801238e-05,
+      "loss": 3.4498,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.299238451141701e-05,
+      "loss": 3.4504,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.298399856390649e-05,
+      "loss": 3.4389,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.297561261639597e-05,
+      "loss": 3.4524,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.296722666888545e-05,
+      "loss": 3.4607,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.295885710017866e-05,
+      "loss": 3.4482,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2950487531471874e-05,
+      "loss": 3.4309,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2942101583961354e-05,
+      "loss": 3.4504,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293371563645083e-05,
+      "loss": 3.4475,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.292532968894031e-05,
+      "loss": 3.4535,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.291694374142979e-05,
+      "loss": 3.4541,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.290855779391927e-05,
+      "loss": 3.4442,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2900171846408754e-05,
+      "loss": 3.4444,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2891785898898234e-05,
+      "loss": 3.4537,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2883416330191443e-05,
+      "loss": 3.4363,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2875030382680923e-05,
+      "loss": 3.4448,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2866644435170403e-05,
+      "loss": 3.4433,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2858274866463615e-05,
+      "loss": 3.4551,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2849888918953095e-05,
+      "loss": 3.4301,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2841502971442575e-05,
+      "loss": 3.4423,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2833117023932055e-05,
+      "loss": 3.4427,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2824747455225264e-05,
+      "loss": 3.4462,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2816361507714744e-05,
+      "loss": 3.4477,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2807975560204224e-05,
+      "loss": 3.4377,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2799589612693707e-05,
+      "loss": 3.4513,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2791220043986917e-05,
+      "loss": 3.4523,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2782834096476397e-05,
+      "loss": 3.4367,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2774448148965877e-05,
+      "loss": 3.4512,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2766062201455356e-05,
+      "loss": 3.4412,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.275769263274857e-05,
+      "loss": 3.45,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274930668523805e-05,
+      "loss": 3.4348,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274092073772753e-05,
+      "loss": 3.4483,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273253479021701e-05,
+      "loss": 3.4359,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2724165221510218e-05,
+      "loss": 3.4496,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2715779273999698e-05,
+      "loss": 3.4422,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2707393326489178e-05,
+      "loss": 3.4504,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.269900737897866e-05,
+      "loss": 3.443,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.269063781027187e-05,
+      "loss": 3.4443,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268225186276135e-05,
+      "loss": 3.4293,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.267386591525083e-05,
+      "loss": 3.4576,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2665479967740307e-05,
+      "loss": 3.4421,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2657110399033523e-05,
+      "loss": 3.451,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2648724451523003e-05,
+      "loss": 3.4361,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2640338504012483e-05,
+      "loss": 3.4531,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.263195255650196e-05,
+      "loss": 3.4488,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262358298779517e-05,
+      "loss": 3.4412,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.261519704028465e-05,
+      "loss": 3.4434,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.260681109277413e-05,
+      "loss": 3.4471,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.259842514526361e-05,
+      "loss": 3.4478,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2590055576556824e-05,
+      "loss": 3.4613,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2581669629046304e-05,
+      "loss": 3.4344,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257328368153578e-05,
+      "loss": 3.4505,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.256489773402526e-05,
+      "loss": 3.4383,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2556528165318473e-05,
+      "loss": 3.4481,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2548142217807956e-05,
+      "loss": 3.4504,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2539756270297433e-05,
+      "loss": 3.4374,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2531370322786913e-05,
+      "loss": 3.454,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2523000754080125e-05,
+      "loss": 3.4405,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2514614806569605e-05,
+      "loss": 3.4521,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9103333950042725,
+      "eval_runtime": 302.4266,
+      "eval_samples_per_second": 1261.764,
+      "eval_steps_per_second": 39.431,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2506228859059085e-05,
+      "loss": 3.4354,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2497842911548565e-05,
+      "loss": 3.4451,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2489456964038045e-05,
+      "loss": 3.4481,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2481071016527525e-05,
+      "loss": 3.4488,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2472701447820734e-05,
+      "loss": 3.4543,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2464315500310214e-05,
+      "loss": 3.442,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2455929552799694e-05,
+      "loss": 3.4332,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2447543605289177e-05,
+      "loss": 3.4381,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2439174036582386e-05,
+      "loss": 3.4477,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2430788089071866e-05,
+      "loss": 3.448,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2422402141561346e-05,
+      "loss": 3.4435,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2414016194050826e-05,
+      "loss": 3.4488,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.240566300414777e-05,
+      "loss": 3.4341,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239727705663725e-05,
+      "loss": 3.4405,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2388891109126728e-05,
+      "loss": 3.44,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2380505161616208e-05,
+      "loss": 3.4262,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2372119214105688e-05,
+      "loss": 3.4418,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2363733266595168e-05,
+      "loss": 3.4346,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2355347319084648e-05,
+      "loss": 3.4329,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234696137157413e-05,
+      "loss": 3.4703,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233859180286734e-05,
+      "loss": 3.442,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233020585535682e-05,
+      "loss": 3.4417,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23218199078463e-05,
+      "loss": 3.4451,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231343396033578e-05,
+      "loss": 3.4498,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2305064391628992e-05,
+      "loss": 3.4265,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2296678444118472e-05,
+      "loss": 3.442,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2288292496607952e-05,
+      "loss": 3.4398,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2279906549097432e-05,
+      "loss": 3.4324,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227153698039064e-05,
+      "loss": 3.4358,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.226315103288012e-05,
+      "loss": 3.4306,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.22547650853696e-05,
+      "loss": 3.441,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.224637913785908e-05,
+      "loss": 3.4457,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2238009569152294e-05,
+      "loss": 3.4413,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2229623621641774e-05,
+      "loss": 3.4393,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2221237674131254e-05,
+      "loss": 3.4414,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2212851726620734e-05,
+      "loss": 3.4469,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2204482157913946e-05,
+      "loss": 3.4332,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2196096210403426e-05,
+      "loss": 3.4515,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2187710262892906e-05,
+      "loss": 3.4274,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2179324315382383e-05,
+      "loss": 3.4443,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2170954746675595e-05,
+      "loss": 3.428,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2162568799165075e-05,
+      "loss": 3.4603,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2154182851654555e-05,
+      "loss": 3.4408,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2145796904144035e-05,
+      "loss": 3.443,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2137427335437247e-05,
+      "loss": 3.4443,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2129041387926727e-05,
+      "loss": 3.442,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2120655440416204e-05,
+      "loss": 3.4394,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2112269492905684e-05,
+      "loss": 3.4305,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2103899924198896e-05,
+      "loss": 3.4186,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.209551397668838e-05,
+      "loss": 3.454,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2087128029177856e-05,
+      "loss": 3.4362,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2078742081667336e-05,
+      "loss": 3.45,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.207037251296055e-05,
+      "loss": 3.432,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.206198656545003e-05,
+      "loss": 3.4384,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205360061793951e-05,
+      "loss": 3.4259,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204521467042899e-05,
+      "loss": 3.4444,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.20368451017222e-05,
+      "loss": 3.443,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2028459154211678e-05,
+      "loss": 3.4475,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2020073206701158e-05,
+      "loss": 3.4462,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2011687259190638e-05,
+      "loss": 3.4287,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200331769048385e-05,
+      "loss": 3.4363,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199493174297333e-05,
+      "loss": 3.4491,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.198654579546281e-05,
+      "loss": 3.428,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197815984795229e-05,
+      "loss": 3.4336,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1969790279245502e-05,
+      "loss": 3.4385,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.196140433173498e-05,
+      "loss": 3.4449,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1953018384224462e-05,
+      "loss": 3.42,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1944648815517675e-05,
+      "loss": 3.4306,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.193626286800715e-05,
+      "loss": 3.4314,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.192787692049663e-05,
+      "loss": 3.4345,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191949097298611e-05,
+      "loss": 3.4425,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1911121404279324e-05,
+      "loss": 3.4381,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1902735456768804e-05,
+      "loss": 3.4435,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1894349509258284e-05,
+      "loss": 3.4288,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1885963561747764e-05,
+      "loss": 3.4557,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1877593993040976e-05,
+      "loss": 3.4281,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1869208045530453e-05,
+      "loss": 3.445,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1860822098019933e-05,
+      "loss": 3.4376,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1852436150509416e-05,
+      "loss": 3.4367,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1844066581802625e-05,
+      "loss": 3.4475,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1835680634292105e-05,
+      "loss": 3.4405,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1827294686781585e-05,
+      "loss": 3.4381,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1818908739271065e-05,
+      "loss": 3.4419,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1810522791760545e-05,
+      "loss": 3.4275,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1802136844250025e-05,
+      "loss": 3.4334,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1793750896739505e-05,
+      "loss": 3.438,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1785381328032717e-05,
+      "loss": 3.4429,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1776995380522197e-05,
+      "loss": 3.4368,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1768609433011677e-05,
+      "loss": 3.4285,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1760223485501157e-05,
+      "loss": 3.4349,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.175185391679437e-05,
+      "loss": 3.4409,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.174346796928385e-05,
+      "loss": 3.4384,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.173508202177333e-05,
+      "loss": 3.4266,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1726696074262806e-05,
+      "loss": 3.4414,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.171832650555602e-05,
+      "loss": 3.4466,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.17099405580455e-05,
+      "loss": 3.4387,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.170157098933871e-05,
+      "loss": 3.4191,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169318504182819e-05,
+      "loss": 3.4391,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.168479909431767e-05,
+      "loss": 3.4299,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.167641314680715e-05,
+      "loss": 3.4446,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.166802719929663e-05,
+      "loss": 3.4424,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1659641251786107e-05,
+      "loss": 3.4323,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1651255304275587e-05,
+      "loss": 3.4321,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.164286935676507e-05,
+      "loss": 3.4432,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.163449978805828e-05,
+      "loss": 3.4245,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.162611384054776e-05,
+      "loss": 3.4358,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.161772789303724e-05,
+      "loss": 3.4291,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1609358324330452e-05,
+      "loss": 3.4486,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1600972376819932e-05,
+      "loss": 3.4175,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1592586429309412e-05,
+      "loss": 3.4294,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1584200481798892e-05,
+      "loss": 3.4303,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1575830913092105e-05,
+      "loss": 3.4393,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.156744496558158e-05,
+      "loss": 3.4315,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.155905901807106e-05,
+      "loss": 3.4278,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.155067307056054e-05,
+      "loss": 3.4375,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1542287123050024e-05,
+      "loss": 3.4416,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1533917554343234e-05,
+      "loss": 3.4227,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1525531606832714e-05,
+      "loss": 3.4428,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1517145659322193e-05,
+      "loss": 3.4303,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1508776090615403e-05,
+      "loss": 3.4359,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1500390143104886e-05,
+      "loss": 3.4253,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1492004195594366e-05,
+      "loss": 3.4345,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1483618248083846e-05,
+      "loss": 3.4257,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1475248679377055e-05,
+      "loss": 3.4358,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1466862731866535e-05,
+      "loss": 3.4363,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1458476784356015e-05,
+      "loss": 3.4342,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1450090836845495e-05,
+      "loss": 3.4324,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1441721268138707e-05,
+      "loss": 3.4333,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1433335320628187e-05,
+      "loss": 3.4208,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1424949373117667e-05,
+      "loss": 3.4474,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1416563425607147e-05,
+      "loss": 3.4332,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1408193856900356e-05,
+      "loss": 3.4363,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.139980790938984e-05,
+      "loss": 3.427,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.139142196187932e-05,
+      "loss": 3.4461,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.13830360143688e-05,
+      "loss": 3.4351,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.137465006685828e-05,
+      "loss": 3.4293,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1366264119347756e-05,
+      "loss": 3.4299,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1357878171837236e-05,
+      "loss": 3.4339,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1349492224326716e-05,
+      "loss": 3.4362,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.134112265561993e-05,
+      "loss": 3.4504,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133273670810941e-05,
+      "loss": 3.4203,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.132435076059889e-05,
+      "loss": 3.4373,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1315964813088368e-05,
+      "loss": 3.4296,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.130759524438158e-05,
+      "loss": 3.439,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1299209296871057e-05,
+      "loss": 3.4349,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.129082334936054e-05,
+      "loss": 3.4304,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.128243740185002e-05,
+      "loss": 3.4419,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127406783314323e-05,
+      "loss": 3.4298,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.126568188563271e-05,
+      "loss": 3.4393,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9114298820495605,
+      "eval_runtime": 302.4484,
+      "eval_samples_per_second": 1261.673,
+      "eval_steps_per_second": 39.428,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.125729593812219e-05,
+      "loss": 3.432,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.124890999061167e-05,
+      "loss": 3.4325,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1240540421904882e-05,
+      "loss": 3.4365,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1232154474394362e-05,
+      "loss": 3.4388,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1223768526883842e-05,
+      "loss": 3.4477,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1215382579373322e-05,
+      "loss": 3.4285,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.120701301066653e-05,
+      "loss": 3.4237,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.119862706315601e-05,
+      "loss": 3.4235,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1190241115645494e-05,
+      "loss": 3.4403,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1181855168134974e-05,
+      "loss": 3.4398,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1173485599428183e-05,
+      "loss": 3.4293,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1165099651917663e-05,
+      "loss": 3.4367,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1156730083210876e-05,
+      "loss": 3.4276,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1148344135700356e-05,
+      "loss": 3.4284,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1139958188189836e-05,
+      "loss": 3.4245,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1131572240679316e-05,
+      "loss": 3.4195,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1123186293168796e-05,
+      "loss": 3.4289,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1114800345658276e-05,
+      "loss": 3.425,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1106414398147756e-05,
+      "loss": 3.4201,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1098028450637236e-05,
+      "loss": 3.4548,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1089658881930448e-05,
+      "loss": 3.4339,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1081272934419928e-05,
+      "loss": 3.4293,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1072886986909408e-05,
+      "loss": 3.4313,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1064501039398884e-05,
+      "loss": 3.4353,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1056131470692097e-05,
+      "loss": 3.4182,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1047745523181577e-05,
+      "loss": 3.4332,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1039359575671057e-05,
+      "loss": 3.4287,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1030973628160537e-05,
+      "loss": 3.4238,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.102260405945375e-05,
+      "loss": 3.4216,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.101421811194323e-05,
+      "loss": 3.4185,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.100583216443271e-05,
+      "loss": 3.4291,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.099746259572592e-05,
+      "loss": 3.4341,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.09890766482154e-05,
+      "loss": 3.4318,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098069070070488e-05,
+      "loss": 3.4267,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0972304753194358e-05,
+      "loss": 3.4294,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.096393518448757e-05,
+      "loss": 3.4325,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.095554923697705e-05,
+      "loss": 3.4276,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.094716328946653e-05,
+      "loss": 3.4331,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.093877734195601e-05,
+      "loss": 3.4191,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.093039139444549e-05,
+      "loss": 3.431,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0922021825738703e-05,
+      "loss": 3.4196,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0913635878228183e-05,
+      "loss": 3.4471,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.090524993071766e-05,
+      "loss": 3.4299,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0896880362010872e-05,
+      "loss": 3.4287,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0888494414500352e-05,
+      "loss": 3.437,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0880108466989832e-05,
+      "loss": 3.4291,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0871722519479312e-05,
+      "loss": 3.4293,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0863352950772524e-05,
+      "loss": 3.4216,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0854967003262004e-05,
+      "loss": 3.4035,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.084658105575148e-05,
+      "loss": 3.4449,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0838195108240964e-05,
+      "loss": 3.4234,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0829825539534177e-05,
+      "loss": 3.4373,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0821439592023653e-05,
+      "loss": 3.4218,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0813053644513133e-05,
+      "loss": 3.4286,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0804667697002613e-05,
+      "loss": 3.4125,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0796298128295826e-05,
+      "loss": 3.4319,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0787912180785306e-05,
+      "loss": 3.4332,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0779526233274786e-05,
+      "loss": 3.4336,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0771140285764266e-05,
+      "loss": 3.4363,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0762770717057478e-05,
+      "loss": 3.4192,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0754384769546955e-05,
+      "loss": 3.4225,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0745998822036435e-05,
+      "loss": 3.4389,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0737612874525918e-05,
+      "loss": 3.4191,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0729243305819127e-05,
+      "loss": 3.4202,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0720857358308607e-05,
+      "loss": 3.4292,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0712471410798087e-05,
+      "loss": 3.43,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0704085463287567e-05,
+      "loss": 3.4133,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.069571589458078e-05,
+      "loss": 3.4203,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.068732994707026e-05,
+      "loss": 3.4161,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.067894399955974e-05,
+      "loss": 3.425,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.067055805204922e-05,
+      "loss": 3.4312,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0662188483342428e-05,
+      "loss": 3.4272,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0653802535831908e-05,
+      "loss": 3.427,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0645416588321388e-05,
+      "loss": 3.4241,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.063703064081087e-05,
+      "loss": 3.4429,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.062866107210408e-05,
+      "loss": 3.4168,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.062027512459356e-05,
+      "loss": 3.4378,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.061188917708304e-05,
+      "loss": 3.4254,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060350322957252e-05,
+      "loss": 3.4201,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0595133660865733e-05,
+      "loss": 3.4391,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0586747713355213e-05,
+      "loss": 3.4324,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0578361765844693e-05,
+      "loss": 3.4282,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0569975818334173e-05,
+      "loss": 3.4314,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0561589870823653e-05,
+      "loss": 3.4185,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0553203923313133e-05,
+      "loss": 3.423,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.054481797580261e-05,
+      "loss": 3.4259,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0536448407095825e-05,
+      "loss": 3.4338,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0528062459585305e-05,
+      "loss": 3.4226,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0519676512074782e-05,
+      "loss": 3.4206,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.051129056456426e-05,
+      "loss": 3.4187,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0502920995857474e-05,
+      "loss": 3.4309,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0494535048346954e-05,
+      "loss": 3.4282,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0486149100836434e-05,
+      "loss": 3.4168,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0477763153325914e-05,
+      "loss": 3.4309,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0469393584619127e-05,
+      "loss": 3.4338,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0461007637108606e-05,
+      "loss": 3.4254,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0452638068401816e-05,
+      "loss": 3.4096,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0444252120891296e-05,
+      "loss": 3.426,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0435866173380775e-05,
+      "loss": 3.4186,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0427480225870255e-05,
+      "loss": 3.4362,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0419094278359735e-05,
+      "loss": 3.4309,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0410708330849215e-05,
+      "loss": 3.4231,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0402322383338695e-05,
+      "loss": 3.422,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0393936435828175e-05,
+      "loss": 3.4266,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0385566867121388e-05,
+      "loss": 3.4175,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0377180919610868e-05,
+      "loss": 3.4254,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0368794972100348e-05,
+      "loss": 3.4144,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0360409024589828e-05,
+      "loss": 3.4368,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0352039455883037e-05,
+      "loss": 3.4085,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0343653508372517e-05,
+      "loss": 3.4172,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0335267560861997e-05,
+      "loss": 3.4175,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.032689799215521e-05,
+      "loss": 3.4275,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.031851204464469e-05,
+      "loss": 3.4178,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.031012609713417e-05,
+      "loss": 3.4203,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.030174014962365e-05,
+      "loss": 3.4256,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0293370580916858e-05,
+      "loss": 3.4317,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.028498463340634e-05,
+      "loss": 3.4145,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.027659868589582e-05,
+      "loss": 3.4287,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.02682127383853e-05,
+      "loss": 3.4243,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025984316967851e-05,
+      "loss": 3.4216,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025145722216799e-05,
+      "loss": 3.4165,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.024307127465747e-05,
+      "loss": 3.4208,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.023468532714695e-05,
+      "loss": 3.4179,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0226315758440163e-05,
+      "loss": 3.4266,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0217929810929643e-05,
+      "loss": 3.426,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0209543863419123e-05,
+      "loss": 3.424,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0201157915908603e-05,
+      "loss": 3.4216,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0192788347201812e-05,
+      "loss": 3.4221,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0184418778495024e-05,
+      "loss": 3.4087,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0176032830984504e-05,
+      "loss": 3.437,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0167646883473984e-05,
+      "loss": 3.4202,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0159260935963464e-05,
+      "loss": 3.4272,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0150874988452944e-05,
+      "loss": 3.4156,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0142489040942424e-05,
+      "loss": 3.4313,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0134103093431904e-05,
+      "loss": 3.4231,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0125717145921384e-05,
+      "loss": 3.418,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0117347577214596e-05,
+      "loss": 3.424,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0108961629704076e-05,
+      "loss": 3.421,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0100575682193556e-05,
+      "loss": 3.4249,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0092189734683033e-05,
+      "loss": 3.4449,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0083820165976245e-05,
+      "loss": 3.4072,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.007543421846573e-05,
+      "loss": 3.4301,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0067048270955205e-05,
+      "loss": 3.4172,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0058662323444685e-05,
+      "loss": 3.4246,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0050292754737898e-05,
+      "loss": 3.4228,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0041906807227378e-05,
+      "loss": 3.4227,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0033520859716858e-05,
+      "loss": 3.4284,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0025134912206338e-05,
+      "loss": 3.4221,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.001676534349955e-05,
+      "loss": 3.4263,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9124810695648193,
+      "eval_runtime": 302.4011,
+      "eval_samples_per_second": 1261.87,
+      "eval_steps_per_second": 39.434,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.000837939598903e-05,
+      "loss": 3.4205,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9999993448478507e-05,
+      "loss": 3.4212,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9991607500967987e-05,
+      "loss": 3.4227,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9983221553457467e-05,
+      "loss": 3.4249,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997485198475068e-05,
+      "loss": 3.435,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996646603724016e-05,
+      "loss": 3.4183,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.995808008972964e-05,
+      "loss": 3.4152,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.994969414221912e-05,
+      "loss": 3.4158,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9941324573512328e-05,
+      "loss": 3.4244,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.993293862600181e-05,
+      "loss": 3.4273,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992455267849129e-05,
+      "loss": 3.4172,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.991616673098077e-05,
+      "loss": 3.4242,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9907813541077713e-05,
+      "loss": 3.4201,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9899427593567193e-05,
+      "loss": 3.4137,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9891041646056673e-05,
+      "loss": 3.4171,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9882655698546153e-05,
+      "loss": 3.4054,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9874269751035633e-05,
+      "loss": 3.4171,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9865883803525113e-05,
+      "loss": 3.4112,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9857497856014593e-05,
+      "loss": 3.4121,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9849111908504073e-05,
+      "loss": 3.4419,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984074233979728e-05,
+      "loss": 3.4267,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9832356392286765e-05,
+      "loss": 3.4201,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9823970444776245e-05,
+      "loss": 3.421,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9815584497265725e-05,
+      "loss": 3.4276,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9807214928558934e-05,
+      "loss": 3.4021,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9798828981048414e-05,
+      "loss": 3.4235,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9790443033537894e-05,
+      "loss": 3.4217,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9782057086027374e-05,
+      "loss": 3.4107,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9773687517320586e-05,
+      "loss": 3.4081,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9765301569810066e-05,
+      "loss": 3.4105,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9756915622299546e-05,
+      "loss": 3.419,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9748529674789026e-05,
+      "loss": 3.425,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9740160106082235e-05,
+      "loss": 3.419,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.973177415857172e-05,
+      "loss": 3.4165,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.97233882110612e-05,
+      "loss": 3.4191,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.971500226355068e-05,
+      "loss": 3.4209,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9706632694843888e-05,
+      "loss": 3.4189,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9698246747333368e-05,
+      "loss": 3.4199,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9689860799822848e-05,
+      "loss": 3.4064,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9681474852312328e-05,
+      "loss": 3.4191,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.967310528360554e-05,
+      "loss": 3.4098,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.966471933609502e-05,
+      "loss": 3.435,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.96563333885845e-05,
+      "loss": 3.4227,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.964794744107398e-05,
+      "loss": 3.4176,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.963957787236719e-05,
+      "loss": 3.425,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.963119192485667e-05,
+      "loss": 3.4173,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9622805977346152e-05,
+      "loss": 3.422,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.961443640863936e-05,
+      "loss": 3.4114,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.960605046112884e-05,
+      "loss": 3.3879,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.959766451361832e-05,
+      "loss": 3.4317,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95892785661078e-05,
+      "loss": 3.4119,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9580908997401014e-05,
+      "loss": 3.4304,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9572523049890494e-05,
+      "loss": 3.4108,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9564137102379974e-05,
+      "loss": 3.4166,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9555751154869454e-05,
+      "loss": 3.404,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9547381586162663e-05,
+      "loss": 3.4197,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9538995638652143e-05,
+      "loss": 3.4257,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9530609691141623e-05,
+      "loss": 3.4201,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9522223743631103e-05,
+      "loss": 3.4295,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9513854174924315e-05,
+      "loss": 3.4115,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9505468227413795e-05,
+      "loss": 3.4099,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9497082279903275e-05,
+      "loss": 3.4275,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.948869633239275e-05,
+      "loss": 3.4099,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9480326763685967e-05,
+      "loss": 3.4121,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9471940816175447e-05,
+      "loss": 3.4153,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9463554868664927e-05,
+      "loss": 3.4221,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9455168921154404e-05,
+      "loss": 3.4016,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9446799352447616e-05,
+      "loss": 3.4114,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9438413404937096e-05,
+      "loss": 3.4063,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9430027457426576e-05,
+      "loss": 3.4122,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9421641509916056e-05,
+      "loss": 3.4221,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.941327194120927e-05,
+      "loss": 3.4149,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.940488599369875e-05,
+      "loss": 3.4174,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9396500046188225e-05,
+      "loss": 3.4111,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9388114098677705e-05,
+      "loss": 3.4346,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937974452997092e-05,
+      "loss": 3.4084,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.93713585824604e-05,
+      "loss": 3.4231,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936298901375361e-05,
+      "loss": 3.4152,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.935460306624309e-05,
+      "loss": 3.409,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.934621711873257e-05,
+      "loss": 3.4276,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.933783117122205e-05,
+      "loss": 3.4212,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.932944522371153e-05,
+      "loss": 3.4183,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.932105927620101e-05,
+      "loss": 3.4163,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.931267332869049e-05,
+      "loss": 3.4111,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.930428738117997e-05,
+      "loss": 3.408,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.929591781247318e-05,
+      "loss": 3.4181,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.928753186496266e-05,
+      "loss": 3.4218,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9279145917452142e-05,
+      "loss": 3.4117,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9270759969941622e-05,
+      "loss": 3.4126,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.926239040123483e-05,
+      "loss": 3.4093,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.925400445372431e-05,
+      "loss": 3.4206,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.924561850621379e-05,
+      "loss": 3.4164,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.923723255870327e-05,
+      "loss": 3.4065,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9228862989996484e-05,
+      "loss": 3.4206,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9220477042485964e-05,
+      "loss": 3.4243,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9212091094975443e-05,
+      "loss": 3.4179,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9203721526268653e-05,
+      "loss": 3.4005,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9195335578758133e-05,
+      "loss": 3.4132,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9186949631247612e-05,
+      "loss": 3.4084,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9178563683737092e-05,
+      "loss": 3.4234,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9170177736226576e-05,
+      "loss": 3.4176,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9161791788716056e-05,
+      "loss": 3.4191,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9153405841205532e-05,
+      "loss": 3.4098,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9145019893695012e-05,
+      "loss": 3.4108,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9136650324988225e-05,
+      "loss": 3.4072,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9128264377477705e-05,
+      "loss": 3.415,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9119878429967185e-05,
+      "loss": 3.4037,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9111492482456665e-05,
+      "loss": 3.4253,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9103122913749877e-05,
+      "loss": 3.4024,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9094736966239354e-05,
+      "loss": 3.402,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9086351018728834e-05,
+      "loss": 3.4112,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9077965071218314e-05,
+      "loss": 3.4106,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.906959550251153e-05,
+      "loss": 3.412,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9061209555001006e-05,
+      "loss": 3.4093,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9052823607490486e-05,
+      "loss": 3.4139,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9044437659979966e-05,
+      "loss": 3.4219,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9036068091273175e-05,
+      "loss": 3.4025,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902768214376266e-05,
+      "loss": 3.4153,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901929619625214e-05,
+      "loss": 3.4127,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901092662754535e-05,
+      "loss": 3.4142,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9002540680034827e-05,
+      "loss": 3.4076,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8994154732524307e-05,
+      "loss": 3.4081,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8985768785013787e-05,
+      "loss": 3.4078,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8977399216307e-05,
+      "loss": 3.415,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896901326879648e-05,
+      "loss": 3.4158,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896062732128596e-05,
+      "loss": 3.411,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895224137377544e-05,
+      "loss": 3.412,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894387180506865e-05,
+      "loss": 3.4098,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.893550223636186e-05,
+      "loss": 3.3995,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8927116288851345e-05,
+      "loss": 3.4243,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8918730341340825e-05,
+      "loss": 3.4077,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89103443938303e-05,
+      "loss": 3.4212,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890195844631978e-05,
+      "loss": 3.4033,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889357249880926e-05,
+      "loss": 3.4225,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.888518655129874e-05,
+      "loss": 3.4155,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.887680060378822e-05,
+      "loss": 3.4076,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8868431035081433e-05,
+      "loss": 3.4121,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8860045087570913e-05,
+      "loss": 3.4095,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8851659140060393e-05,
+      "loss": 3.416,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8843273192549873e-05,
+      "loss": 3.4309,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8834903623843082e-05,
+      "loss": 3.3993,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8826517676332566e-05,
+      "loss": 3.4228,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8818131728822046e-05,
+      "loss": 3.4039,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8809745781311526e-05,
+      "loss": 3.4143,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8801376212604735e-05,
+      "loss": 3.414,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8792990265094215e-05,
+      "loss": 3.4113,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8784604317583695e-05,
+      "loss": 3.4165,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8776218370073175e-05,
+      "loss": 3.4148,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8767848801366387e-05,
+      "loss": 3.4128,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9143621921539307,
+      "eval_runtime": 308.5376,
+      "eval_samples_per_second": 1236.773,
+      "eval_steps_per_second": 38.65,
+      "step": 1908000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8759462853855867e-05,
+      "loss": 3.409,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8751076906345347e-05,
+      "loss": 3.4122,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8742690958834827e-05,
+      "loss": 3.4121,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8734321390128036e-05,
+      "loss": 3.4135,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8725935442617516e-05,
+      "loss": 3.4229,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8717549495107e-05,
+      "loss": 3.4072,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.870917992640021e-05,
+      "loss": 3.4056,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.870079397888969e-05,
+      "loss": 3.4073,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.869240803137917e-05,
+      "loss": 3.4122,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8684022083868648e-05,
+      "loss": 3.4172,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.867565251516186e-05,
+      "loss": 3.4109,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.866726656765134e-05,
+      "loss": 3.4113,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.865889699894455e-05,
+      "loss": 3.4116,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.865051105143403e-05,
+      "loss": 3.4012,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.864212510392351e-05,
+      "loss": 3.406,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.863373915641299e-05,
+      "loss": 3.396,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.862535320890247e-05,
+      "loss": 3.4024,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8616967261391953e-05,
+      "loss": 3.403,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.860858131388143e-05,
+      "loss": 3.4012,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.860019536637091e-05,
+      "loss": 3.4234,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8591825797664122e-05,
+      "loss": 3.4163,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.85834398501536e-05,
+      "loss": 3.4115,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8575053902643082e-05,
+      "loss": 3.4108,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8566667955132562e-05,
+      "loss": 3.4154,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8558298386425774e-05,
+      "loss": 3.3945,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.854991243891525e-05,
+      "loss": 3.4108,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.854152649140473e-05,
+      "loss": 3.4104,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853314054389421e-05,
+      "loss": 3.3962,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8524770975187423e-05,
+      "loss": 3.4038,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8516385027676903e-05,
+      "loss": 3.4006,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8507999080166383e-05,
+      "loss": 3.408,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8499613132655863e-05,
+      "loss": 3.4067,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8491243563949072e-05,
+      "loss": 3.4116,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8482857616438552e-05,
+      "loss": 3.4137,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8474471668928036e-05,
+      "loss": 3.4046,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8466085721417516e-05,
+      "loss": 3.4111,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8457716152710725e-05,
+      "loss": 3.4146,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8449330205200205e-05,
+      "loss": 3.4025,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8440944257689685e-05,
+      "loss": 3.3985,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8432558310179165e-05,
+      "loss": 3.4082,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8424188741472377e-05,
+      "loss": 3.4005,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8415802793961857e-05,
+      "loss": 3.4216,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8407416846451337e-05,
+      "loss": 3.4149,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8399030898940817e-05,
+      "loss": 3.4041,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8390661330234026e-05,
+      "loss": 3.4135,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8382275382723506e-05,
+      "loss": 3.4055,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8373889435212986e-05,
+      "loss": 3.4097,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.83655198665062e-05,
+      "loss": 3.409,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.835713391899568e-05,
+      "loss": 3.375,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8348747971485158e-05,
+      "loss": 3.422,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8340362023974638e-05,
+      "loss": 3.4051,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.833199245526785e-05,
+      "loss": 3.4122,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.832360650775733e-05,
+      "loss": 3.4034,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.831522056024681e-05,
+      "loss": 3.4047,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.830683461273629e-05,
+      "loss": 3.3916,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.82984650440295e-05,
+      "loss": 3.4088,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.829007909651898e-05,
+      "loss": 3.4123,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.828169314900846e-05,
+      "loss": 3.4088,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.827330720149794e-05,
+      "loss": 3.4243,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8264937632791152e-05,
+      "loss": 3.4021,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8256551685280632e-05,
+      "loss": 3.3955,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.824818211657384e-05,
+      "loss": 3.4191,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.823979616906332e-05,
+      "loss": 3.3993,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.82314102215528e-05,
+      "loss": 3.3951,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8223024274042284e-05,
+      "loss": 3.4084,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8214638326531764e-05,
+      "loss": 3.4085,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8206252379021244e-05,
+      "loss": 3.3946,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8197866431510724e-05,
+      "loss": 3.3972,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.81894804840002e-05,
+      "loss": 3.3957,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8181110915293413e-05,
+      "loss": 3.3991,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8172724967782893e-05,
+      "loss": 3.4122,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8164339020272377e-05,
+      "loss": 3.4053,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8155953072761853e-05,
+      "loss": 3.4074,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8147583504055066e-05,
+      "loss": 3.3993,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8139197556544546e-05,
+      "loss": 3.422,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8130811609034022e-05,
+      "loss": 3.3989,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8122425661523505e-05,
+      "loss": 3.4164,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8114056092816718e-05,
+      "loss": 3.4036,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8105670145306198e-05,
+      "loss": 3.4003,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8097300576599407e-05,
+      "loss": 3.4127,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8088914629088887e-05,
+      "loss": 3.4134,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8080528681578367e-05,
+      "loss": 3.4063,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8072142734067847e-05,
+      "loss": 3.4087,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8063756786557327e-05,
+      "loss": 3.4014,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8055370839046807e-05,
+      "loss": 3.3952,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8046984891536287e-05,
+      "loss": 3.4064,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8038598944025767e-05,
+      "loss": 3.4106,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8030229375318976e-05,
+      "loss": 3.4005,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.802184342780846e-05,
+      "loss": 3.4025,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.801345748029794e-05,
+      "loss": 3.3987,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.800507153278742e-05,
+      "loss": 3.4067,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7996701964080628e-05,
+      "loss": 3.4073,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7988316016570108e-05,
+      "loss": 3.3965,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7979930069059588e-05,
+      "loss": 3.4115,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.79715605003528e-05,
+      "loss": 3.4072,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796317455284228e-05,
+      "loss": 3.4091,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7954804984135493e-05,
+      "loss": 3.3884,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.794641903662497e-05,
+      "loss": 3.4018,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.793803308911445e-05,
+      "loss": 3.3999,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.792964714160393e-05,
+      "loss": 3.4146,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.792126119409341e-05,
+      "loss": 3.408,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7912875246582893e-05,
+      "loss": 3.4119,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7904489299072373e-05,
+      "loss": 3.4016,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7896103351561853e-05,
+      "loss": 3.3998,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7887733782855062e-05,
+      "loss": 3.3951,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7879347835344542e-05,
+      "loss": 3.4052,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.787096188783402e-05,
+      "loss": 3.3968,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.78625759403235e-05,
+      "loss": 3.4075,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7854206371616714e-05,
+      "loss": 3.3977,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7845820424106194e-05,
+      "loss": 3.3906,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7837434476595674e-05,
+      "loss": 3.3993,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.782904852908515e-05,
+      "loss": 3.4005,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7820678960378363e-05,
+      "loss": 3.3987,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7812293012867846e-05,
+      "loss": 3.3982,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7803907065357326e-05,
+      "loss": 3.4085,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7795521117846803e-05,
+      "loss": 3.4108,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7787151549140015e-05,
+      "loss": 3.3941,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7778765601629495e-05,
+      "loss": 3.4019,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7770379654118975e-05,
+      "loss": 3.4042,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7761993706608455e-05,
+      "loss": 3.4004,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7753624137901668e-05,
+      "loss": 3.3998,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7745238190391148e-05,
+      "loss": 3.395,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7736852242880624e-05,
+      "loss": 3.4003,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7728466295370104e-05,
+      "loss": 3.401,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7720096726663317e-05,
+      "loss": 3.4066,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.77117107791528e-05,
+      "loss": 3.4016,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7703324831642277e-05,
+      "loss": 3.3977,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7694938884131757e-05,
+      "loss": 3.401,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7686585694228698e-05,
+      "loss": 3.3868,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7678199746718178e-05,
+      "loss": 3.4124,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.766981379920766e-05,
+      "loss": 3.3961,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.766142785169714e-05,
+      "loss": 3.4144,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.765304190418662e-05,
+      "loss": 3.3916,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7644655956676098e-05,
+      "loss": 3.4094,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7636270009165578e-05,
+      "loss": 3.4077,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7627884061655058e-05,
+      "loss": 3.3959,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761951449294827e-05,
+      "loss": 3.3989,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761112854543775e-05,
+      "loss": 3.3956,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.760274259792723e-05,
+      "loss": 3.4102,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759435665041671e-05,
+      "loss": 3.4158,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.758598708170992e-05,
+      "loss": 3.3909,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.75776011341994e-05,
+      "loss": 3.4073,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7569215186688883e-05,
+      "loss": 3.398,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7560829239178363e-05,
+      "loss": 3.4043,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7552459670471572e-05,
+      "loss": 3.4006,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754407372296105e-05,
+      "loss": 3.4056,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.753568777545053e-05,
+      "loss": 3.402,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.752730182794001e-05,
+      "loss": 3.404,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7518932259233224e-05,
+      "loss": 3.4044,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9172556400299072,
+      "eval_runtime": 310.8268,
+      "eval_samples_per_second": 1227.664,
+      "eval_steps_per_second": 38.365,
+      "step": 1984320
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.287473272705751e+18,
+  "trial_name": null,
+  "trial_params": null
+}