diff --git "a/checkpoint-1679040/trainer_state.json" "b/checkpoint-1679040/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1679040/trainer_state.json"
@@ -0,0 +1,19875 @@
+{
+  "best_metric": 4.043795108795166,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/lstm/4/checkpoints/checkpoint-1679040",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1679040,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8195,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5627,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0613,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9898,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9202,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.7783,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.655,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.5569,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.4699,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.398,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3437,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.2739,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2187,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1474,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.0987,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0334,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986585759743914e-05,
+      "loss": 6.0066,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985747164992862e-05,
+      "loss": 5.9522,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.9096,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.8792,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.8584,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8146,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.7927,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7571,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7259,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7099,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6728,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.6622,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6369,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6098,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5975,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5742,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5614,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5445,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5246,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5164,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5095,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4711,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4617,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4483,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4321,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4199,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4204,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4004,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3849,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3784,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3678,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3617,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3377,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3256,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3342,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3236,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.31,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2879,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2967,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2745,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2905,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2562,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2603,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2399,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2296,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.232,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2168,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2111,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2033,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2122,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.1851,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1755,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1604,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.16,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.174,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1638,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1459,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1416,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1417,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1357,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1136,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1178,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1047,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1061,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.101,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0895,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0926,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0819,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0666,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0658,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 5.0708,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 5.0638,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 5.0512,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 5.0381,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.0418,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0302,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0405,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.019,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0136,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0313,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.9991,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9929,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9976,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9954,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9961,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9851,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9779,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.9699,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9699,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.954,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9537,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9583,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9465,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9388,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9357,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9266,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9353,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9216,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9178,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9199,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9133,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9077,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9067,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9049,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9038,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8839,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8958,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8874,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.8875,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8681,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8786,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8627,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8693,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8456,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8694,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8726,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8531,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8403,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8512,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8454,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8435,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.837,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8411,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8312,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.8254,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8098,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8225,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8202,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8087,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8218,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8005,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.8099,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.8103,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7988,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.811893463134766,
+      "eval_runtime": 292.9712,
+      "eval_samples_per_second": 1302.486,
+      "eval_steps_per_second": 40.704,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.7876,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.7943,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.7965,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7868,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7874,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7719,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7764,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7665,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7725,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7632,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7691,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.7712,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.7568,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.7552,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.7383,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7496,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.7432,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7373,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.7345,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.763,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.7341,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7351,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.7393,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7291,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7254,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7136,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7218,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7193,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7051,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7141,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7043,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7141,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.7124,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.698,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.7069,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7145,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6838,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6945,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6912,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6821,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6827,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6928,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6851,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6797,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6818,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.6762,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6795,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6619,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6589,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6755,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6735,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6629,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6543,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6628,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6448,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.671,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.651,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6617,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.6443,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6399,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823958979613631e-05,
+      "loss": 4.6419,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.6478,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6307,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6353,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.649,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6293,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6194,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6202,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6178,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.64,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.6351,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6184,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.6231,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6266,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6238,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6074,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6198,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097094203672393e-05,
+      "loss": 4.6104,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8088708256161873e-05,
+      "loss": 4.615,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808032230865135e-05,
+      "loss": 4.607,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.6076,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6094,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6113,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.5931,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.6022,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.6004,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6044,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5974,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.5855,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.5939,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5867,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5986,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.5782,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5829,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5951,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5845,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5656,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5778,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5824,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.583,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5744,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5702,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.566,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.5728,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.5536,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5606,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5671,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5655,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.549,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5597,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5473,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.5633,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5503,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.5422,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5593,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5533,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.5464,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5496,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5473,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5522,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5315,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5515,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.539,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5432,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.5333,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.5399,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686330184890494e-05,
+      "loss": 4.5279,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76779606161837e-05,
+      "loss": 4.5356,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766957466867318e-05,
+      "loss": 4.5144,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766118872116266e-05,
+      "loss": 4.5448,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652802773652136e-05,
+      "loss": 4.5403,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644416826141616e-05,
+      "loss": 4.5321,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636030878631096e-05,
+      "loss": 4.5203,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.5338,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.5334,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5244,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760250346739275e-05,
+      "loss": 4.5194,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759411751988223e-05,
+      "loss": 4.5292,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.522,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.5287,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.5059,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5218,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5196,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5113,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.5258,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5057,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751866037109128e-05,
+      "loss": 4.5154,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751027442358077e-05,
+      "loss": 4.5135,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.5128,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.5289154052734375,
+      "eval_runtime": 296.6014,
+      "eval_samples_per_second": 1286.545,
+      "eval_steps_per_second": 40.205,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.5015,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.5077,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5137,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.5066,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.5085,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.4906,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.5,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.4921,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.5013,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.4946,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.5021,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.5031,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.4863,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.4911,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4744,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4862,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4858,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.479,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4782,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5053,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4835,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4809,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4885,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4837,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4768,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4678,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275500650893664e-05,
+      "loss": 4.4774,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726713108218687e-05,
+      "loss": 4.4712,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725874513467635e-05,
+      "loss": 4.4651,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725035918716583e-05,
+      "loss": 4.4735,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.4637,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723358729214479e-05,
+      "loss": 4.4767,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.4713,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721683177592748e-05,
+      "loss": 4.4609,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720844582841696e-05,
+      "loss": 4.4723,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720005988090644e-05,
+      "loss": 4.4836,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4495,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4683,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.4577,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.4585,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4532,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714976057464706e-05,
+      "loss": 4.4693,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4619,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4572,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4584,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.4507,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4617,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4421,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091075320877147e-05,
+      "loss": 4.443,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082689373366627e-05,
+      "loss": 4.4556,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.455,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4462,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70575479096388e-05,
+      "loss": 4.4437,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704916196212828e-05,
+      "loss": 4.4515,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4335,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703240644591097e-05,
+      "loss": 4.4582,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702402049840045e-05,
+      "loss": 4.4424,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015634550889924e-05,
+      "loss": 4.4515,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4439,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4352,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4338,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982107139651573e-05,
+      "loss": 4.4444,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697372119214106e-05,
+      "loss": 4.4235,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4337,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.444,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4335,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.419,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4203,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692340550707794e-05,
+      "loss": 4.4196,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691505231717488e-05,
+      "loss": 4.4459,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690666636966436e-05,
+      "loss": 4.4355,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689828042215384e-05,
+      "loss": 4.4253,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4332,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4288,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4333,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864753010915494e-05,
+      "loss": 4.4158,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68563834422087e-05,
+      "loss": 4.4298,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684799749469818e-05,
+      "loss": 4.4235,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683961154718766e-05,
+      "loss": 4.4228,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683122559967714e-05,
+      "loss": 4.4178,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4219,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4225,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4244,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679769818843879e-05,
+      "loss": 4.4124,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678931224092827e-05,
+      "loss": 4.4166,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678092629341775e-05,
+      "loss": 4.4152,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677254034590723e-05,
+      "loss": 4.4195,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.4193,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.402,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.413,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4102,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4189,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.3989,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.4081,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4185,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.4074,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668869724960576e-05,
+      "loss": 4.3915,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.4037,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667192535458472e-05,
+      "loss": 4.4096,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635394070742e-05,
+      "loss": 4.4091,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665515345956368e-05,
+      "loss": 4.4047,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.3964,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.3967,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.4005,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621626048325336e-05,
+      "loss": 4.3866,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613240100814816e-05,
+      "loss": 4.3934,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3981,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.3974,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.3783,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.3957,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3821,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.3945,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.39,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.3804,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65377993308276e-05,
+      "loss": 4.3951,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.394,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.3824,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.3898,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.3882,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649588597207873e-05,
+      "loss": 4.3893,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648750002456821e-05,
+      "loss": 4.3758,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.3891,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64707445083509e-05,
+      "loss": 4.3839,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646235856084038e-05,
+      "loss": 4.3845,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.3797,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3873,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3675,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.3807,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642042882328778e-05,
+      "loss": 4.3614,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641204287577726e-05,
+      "loss": 4.3919,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403673307070474e-05,
+      "loss": 4.3826,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639528735955995e-05,
+      "loss": 4.3794,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638690141204943e-05,
+      "loss": 4.3683,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637851546453891e-05,
+      "loss": 4.3827,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.385,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3732,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.3706,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634500443210429e-05,
+      "loss": 4.3794,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633661848459377e-05,
+      "loss": 4.373,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3798,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3613,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3736,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3707,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3639,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628630279953065e-05,
+      "loss": 4.3823,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.3588,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3686,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.3718,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62527753882923e-05,
+      "loss": 4.3661,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.393002033233643,
+      "eval_runtime": 290.8721,
+      "eval_samples_per_second": 1311.886,
+      "eval_steps_per_second": 40.997,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3654,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3643,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3732,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3655,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.3668,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3542,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.362,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3454,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177334618305085e-05,
+      "loss": 4.3692,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168948670794565e-05,
+      "loss": 4.3525,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3638,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3655,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3535,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3485,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.343,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3448,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3548,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.3434,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3464,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3695,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3519,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3501,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3548,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605157816325474e-05,
+      "loss": 4.3514,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604320859454796e-05,
+      "loss": 4.3465,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603482264703743e-05,
+      "loss": 4.336,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602643669952691e-05,
+      "loss": 4.3524,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.334,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600968118330961e-05,
+      "loss": 4.3385,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600129523579909e-05,
+      "loss": 4.3404,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599290928828857e-05,
+      "loss": 4.3353,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3512,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3428,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596778420336447e-05,
+      "loss": 4.3351,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3448,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595101230834343e-05,
+      "loss": 4.3531,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942626360832906e-05,
+      "loss": 4.3304,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934240413322386e-05,
+      "loss": 4.3413,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3267,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3424,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3267,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3458,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589232705457352e-05,
+      "loss": 4.3374,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883941107063e-05,
+      "loss": 4.3333,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587555515955248e-05,
+      "loss": 4.3369,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3298,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3402,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3241,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.32,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3363,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3313,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3313,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3189,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3313,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.316,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3359,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.323,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3346,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.328,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3202,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3166,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3281,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572470637718551e-05,
+      "loss": 4.3101,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571632042967499e-05,
+      "loss": 4.3175,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570793448216447e-05,
+      "loss": 4.3238,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569954853465395e-05,
+      "loss": 4.3232,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569116258714343e-05,
+      "loss": 4.3013,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568277663963291e-05,
+      "loss": 4.3073,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567439069212239e-05,
+      "loss": 4.3039,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566600474461187e-05,
+      "loss": 4.333,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565761879710135e-05,
+      "loss": 4.3194,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564924922839456e-05,
+      "loss": 4.3153,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564086328088404e-05,
+      "loss": 4.3198,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563247733337352e-05,
+      "loss": 4.3143,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624091385863e-05,
+      "loss": 4.3271,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615721817156206e-05,
+      "loss": 4.3009,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560733586964569e-05,
+      "loss": 4.3219,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559894992213517e-05,
+      "loss": 4.3109,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559056397462465e-05,
+      "loss": 4.3123,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558217802711413e-05,
+      "loss": 4.3141,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557379207960361e-05,
+      "loss": 4.3075,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556540613209309e-05,
+      "loss": 4.3171,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557020184582566e-05,
+      "loss": 4.3132,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554865061587578e-05,
+      "loss": 4.3027,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554026466836526e-05,
+      "loss": 4.3109,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553187872085474e-05,
+      "loss": 4.302,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523492773344215e-05,
+      "loss": 4.3157,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551512320463743e-05,
+      "loss": 4.309,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550673725712691e-05,
+      "loss": 4.2936,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549836768842013e-05,
+      "loss": 4.3085,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548998174090961e-05,
+      "loss": 4.3001,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548159579339909e-05,
+      "loss": 4.31,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547320984588857e-05,
+      "loss": 4.2953,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546482389837804e-05,
+      "loss": 4.3039,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545643795086752e-05,
+      "loss": 4.3135,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448052003357e-05,
+      "loss": 4.2989,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543966605584648e-05,
+      "loss": 4.2915,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543128010833596e-05,
+      "loss": 4.2975,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542289416082544e-05,
+      "loss": 4.2979,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541452459211865e-05,
+      "loss": 4.3109,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540613864460813e-05,
+      "loss": 4.2984,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539775269709761e-05,
+      "loss": 4.2934,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389366749587095e-05,
+      "loss": 4.2932,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380997180880304e-05,
+      "loss": 4.3014,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372611233369784e-05,
+      "loss": 4.2794,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364225285859264e-05,
+      "loss": 4.294,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355855717152473e-05,
+      "loss": 4.294,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347469769641953e-05,
+      "loss": 4.2995,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533908382213143e-05,
+      "loss": 4.2748,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533069787462091e-05,
+      "loss": 4.295,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532231192711039e-05,
+      "loss": 4.2874,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531392597959987e-05,
+      "loss": 4.2856,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530554003208935e-05,
+      "loss": 4.2929,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529715408457883e-05,
+      "loss": 4.2782,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.2984,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528041494716526e-05,
+      "loss": 4.2923,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527202899965474e-05,
+      "loss": 4.2853,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526364305214422e-05,
+      "loss": 4.2929,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52552571046337e-05,
+      "loss": 4.2849,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524687115712318e-05,
+      "loss": 4.2932,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523848520961266e-05,
+      "loss": 4.2733,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523011564090587e-05,
+      "loss": 4.2971,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522172969339535e-05,
+      "loss": 4.2833,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521334374588483e-05,
+      "loss": 4.2854,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520495779837431e-05,
+      "loss": 4.2827,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519657185086379e-05,
+      "loss": 4.2918,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518818590335327e-05,
+      "loss": 4.2764,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517979995584275e-05,
+      "loss": 4.2794,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517143038713596e-05,
+      "loss": 4.2692,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516304443962544e-05,
+      "loss": 4.2936,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515465849211492e-05,
+      "loss": 4.2875,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51462725446044e-05,
+      "loss": 4.2855,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137886597093876e-05,
+      "loss": 4.2737,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129500649583356e-05,
+      "loss": 4.2879,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121114702072836e-05,
+      "loss": 4.2873,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112728754562316e-05,
+      "loss": 4.279,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104342807051795e-05,
+      "loss": 4.2753,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095956859541275e-05,
+      "loss": 4.2856,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087587290834485e-05,
+      "loss": 4.2835,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507920134332397e-05,
+      "loss": 4.285,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507081539581345e-05,
+      "loss": 4.2687,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506242944830293e-05,
+      "loss": 4.2804,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505405987959614e-05,
+      "loss": 4.2771,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504567393208562e-05,
+      "loss": 4.276,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50372879845751e-05,
+      "loss": 4.2849,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502891841586831e-05,
+      "loss": 4.2695,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502053246835779e-05,
+      "loss": 4.2778,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501214652084727e-05,
+      "loss": 4.2745,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500376057333675e-05,
+      "loss": 4.2793,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.310644626617432,
+      "eval_runtime": 288.9659,
+      "eval_samples_per_second": 1320.54,
+      "eval_steps_per_second": 41.268,
+      "step": 305280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.499537462582623e-05,
+      "loss": 4.2742,
+      "step": 305664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.498698867831571e-05,
+      "loss": 4.2749,
+      "step": 306176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.497860273080519e-05,
+      "loss": 4.2825,
+      "step": 306688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.497021678329467e-05,
+      "loss": 4.2731,
+      "step": 307200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4961847214587885e-05,
+      "loss": 4.2807,
+      "step": 307712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4953461267077365e-05,
+      "loss": 4.2633,
+      "step": 308224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4945075319566845e-05,
+      "loss": 4.2713,
+      "step": 308736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4936689372056325e-05,
+      "loss": 4.2582,
+      "step": 309248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4928319803349534e-05,
+      "loss": 4.277,
+      "step": 309760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.491995023464274e-05,
+      "loss": 4.2708,
+      "step": 310272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.491156428713222e-05,
+      "loss": 4.2729,
+      "step": 310784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.49031783396217e-05,
+      "loss": 4.2733,
+      "step": 311296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.489480877091491e-05,
+      "loss": 4.2654,
+      "step": 311808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.488642282340439e-05,
+      "loss": 4.2607,
+      "step": 312320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.487803687589387e-05,
+      "loss": 4.262,
+      "step": 312832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.486965092838336e-05,
+      "loss": 4.251,
+      "step": 313344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.486126498087284e-05,
+      "loss": 4.2707,
+      "step": 313856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.485287903336232e-05,
+      "loss": 4.2566,
+      "step": 314368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.48444930858518e-05,
+      "loss": 4.2579,
+      "step": 314880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.483610713834128e-05,
+      "loss": 4.2863,
+      "step": 315392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.482772119083076e-05,
+      "loss": 4.2601,
+      "step": 315904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.481933524332024e-05,
+      "loss": 4.2654,
+      "step": 316416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.481094929580971e-05,
+      "loss": 4.2679,
+      "step": 316928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.480256334829919e-05,
+      "loss": 4.2632,
+      "step": 317440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.479419377959241e-05,
+      "loss": 4.2618,
+      "step": 317952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.478580783208188e-05,
+      "loss": 4.2529,
+      "step": 318464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4777438263375096e-05,
+      "loss": 4.2676,
+      "step": 318976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4769052315864576e-05,
+      "loss": 4.2442,
+      "step": 319488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4760666368354056e-05,
+      "loss": 4.2561,
+      "step": 320000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4752280420843536e-05,
+      "loss": 4.252,
+      "step": 320512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4743894473333016e-05,
+      "loss": 4.2532,
+      "step": 321024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4735508525822496e-05,
+      "loss": 4.2666,
+      "step": 321536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.472713895711571e-05,
+      "loss": 4.26,
+      "step": 322048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4718753009605185e-05,
+      "loss": 4.2545,
+      "step": 322560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4710367062094665e-05,
+      "loss": 4.2605,
+      "step": 323072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4701981114584145e-05,
+      "loss": 4.2668,
+      "step": 323584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4693611545877354e-05,
+      "loss": 4.2496,
+      "step": 324096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4685225598366834e-05,
+      "loss": 4.2597,
+      "step": 324608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4676839650856314e-05,
+      "loss": 4.2448,
+      "step": 325120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4668453703345794e-05,
+      "loss": 4.26,
+      "step": 325632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.466006775583528e-05,
+      "loss": 4.24,
+      "step": 326144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.465168180832476e-05,
+      "loss": 4.2654,
+      "step": 326656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.464329586081424e-05,
+      "loss": 4.2557,
+      "step": 327168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.463490991330372e-05,
+      "loss": 4.2511,
+      "step": 327680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.462654034459693e-05,
+      "loss": 4.2549,
+      "step": 328192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.461815439708641e-05,
+      "loss": 4.2489,
+      "step": 328704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.460976844957589e-05,
+      "loss": 4.2586,
+      "step": 329216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.460138250206537e-05,
+      "loss": 4.2435,
+      "step": 329728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.459301293335858e-05,
+      "loss": 4.2393,
+      "step": 330240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.458462698584806e-05,
+      "loss": 4.2581,
+      "step": 330752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.457624103833754e-05,
+      "loss": 4.2494,
+      "step": 331264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.456787146963075e-05,
+      "loss": 4.2501,
+      "step": 331776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4559485522120234e-05,
+      "loss": 4.2417,
+      "step": 332288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4551115953413444e-05,
+      "loss": 4.2486,
+      "step": 332800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4542730005902923e-05,
+      "loss": 4.2412,
+      "step": 333312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4534344058392403e-05,
+      "loss": 4.2531,
+      "step": 333824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4525958110881883e-05,
+      "loss": 4.2423,
+      "step": 334336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.451757216337136e-05,
+      "loss": 4.2567,
+      "step": 334848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.450918621586084e-05,
+      "loss": 4.2491,
+      "step": 335360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.450080026835032e-05,
+      "loss": 4.2397,
+      "step": 335872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.44924143208398e-05,
+      "loss": 4.2396,
+      "step": 336384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.448402837332928e-05,
+      "loss": 4.2483,
+      "step": 336896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.447564242581876e-05,
+      "loss": 4.2323,
+      "step": 337408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.446725647830824e-05,
+      "loss": 4.2435,
+      "step": 337920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.445888690960145e-05,
+      "loss": 4.2438,
+      "step": 338432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.445050096209093e-05,
+      "loss": 4.2492,
+      "step": 338944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.444211501458041e-05,
+      "loss": 4.2257,
+      "step": 339456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.44337290670699e-05,
+      "loss": 4.2324,
+      "step": 339968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.442535949836311e-05,
+      "loss": 4.2266,
+      "step": 340480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.441697355085259e-05,
+      "loss": 4.2536,
+      "step": 340992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.440858760334207e-05,
+      "loss": 4.2407,
+      "step": 341504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.440020165583154e-05,
+      "loss": 4.2407,
+      "step": 342016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.439181570832102e-05,
+      "loss": 4.2413,
+      "step": 342528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.43834297608105e-05,
+      "loss": 4.2385,
+      "step": 343040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.437504381329998e-05,
+      "loss": 4.2566,
+      "step": 343552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.436667424459319e-05,
+      "loss": 4.2232,
+      "step": 344064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2437,
+      "step": 344576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2416,
+      "step": 345088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2369,
+      "step": 345600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4333146833354846e-05,
+      "loss": 4.2391,
+      "step": 346112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4324760885844326e-05,
+      "loss": 4.2341,
+      "step": 346624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4316374938333806e-05,
+      "loss": 4.2392,
+      "step": 347136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4307988990823286e-05,
+      "loss": 4.2414,
+      "step": 347648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4299619422116495e-05,
+      "loss": 4.227,
+      "step": 348160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4291233474605975e-05,
+      "loss": 4.2385,
+      "step": 348672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4282847527095455e-05,
+      "loss": 4.2289,
+      "step": 349184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4274461579584935e-05,
+      "loss": 4.241,
+      "step": 349696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4266075632074415e-05,
+      "loss": 4.2342,
+      "step": 350208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4257706063367624e-05,
+      "loss": 4.2226,
+      "step": 350720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4249320115857104e-05,
+      "loss": 4.2346,
+      "step": 351232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.424093416834659e-05,
+      "loss": 4.2275,
+      "step": 351744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.423254822083607e-05,
+      "loss": 4.2369,
+      "step": 352256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.422416227332555e-05,
+      "loss": 4.2201,
+      "step": 352768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.421577632581503e-05,
+      "loss": 4.2332,
+      "step": 353280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.420739037830451e-05,
+      "loss": 4.2453,
+      "step": 353792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.419902080959772e-05,
+      "loss": 4.2253,
+      "step": 354304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.41906348620872e-05,
+      "loss": 4.2147,
+      "step": 354816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.418224891457668e-05,
+      "loss": 4.2254,
+      "step": 355328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.417386296706616e-05,
+      "loss": 4.2294,
+      "step": 355840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.416547701955564e-05,
+      "loss": 4.2352,
+      "step": 356352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.415709107204512e-05,
+      "loss": 4.2302,
+      "step": 356864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.41487051245346e-05,
+      "loss": 4.2216,
+      "step": 357376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.414033555582781e-05,
+      "loss": 4.2207,
+      "step": 357888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.413194960831729e-05,
+      "loss": 4.2314,
+      "step": 358400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4123563660806775e-05,
+      "loss": 4.2078,
+      "step": 358912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4115177713296255e-05,
+      "loss": 4.2215,
+      "step": 359424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4106791765785735e-05,
+      "loss": 4.2232,
+      "step": 359936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.409840581827521e-05,
+      "loss": 4.2285,
+      "step": 360448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.2058,
+      "step": 360960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2252,
+      "step": 361472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.407326435454738e-05,
+      "loss": 4.2171,
+      "step": 361984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.406487840703686e-05,
+      "loss": 4.2187,
+      "step": 362496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.405649245952634e-05,
+      "loss": 4.2189,
+      "step": 363008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.404810651201582e-05,
+      "loss": 4.211,
+      "step": 363520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4039736943309026e-05,
+      "loss": 4.2245,
+      "step": 364032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.403135099579851e-05,
+      "loss": 4.2263,
+      "step": 364544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.402296504828799e-05,
+      "loss": 4.2145,
+      "step": 365056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.401457910077747e-05,
+      "loss": 4.226,
+      "step": 365568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.400619315326695e-05,
+      "loss": 4.2141,
+      "step": 366080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.399780720575643e-05,
+      "loss": 4.2262,
+      "step": 366592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.2081,
+      "step": 367104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.398105168953912e-05,
+      "loss": 4.2287,
+      "step": 367616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.39726657420286e-05,
+      "loss": 4.2089,
+      "step": 368128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.396427979451808e-05,
+      "loss": 4.2201,
+      "step": 368640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.395589384700756e-05,
+      "loss": 4.211,
+      "step": 369152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.394750789949704e-05,
+      "loss": 4.2249,
+      "step": 369664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.393912195198652e-05,
+      "loss": 4.212,
+      "step": 370176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3930736004476e-05,
+      "loss": 4.2142,
+      "step": 370688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.1992,
+      "step": 371200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.39139804882587e-05,
+      "loss": 4.2296,
+      "step": 371712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.2153,
+      "step": 372224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.389720859323766e-05,
+      "loss": 4.2173,
+      "step": 372736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.388882264572714e-05,
+      "loss": 4.2073,
+      "step": 373248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2215,
+      "step": 373760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2236,
+      "step": 374272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.211,
+      "step": 374784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2088,
+      "step": 375296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.2188,
+      "step": 375808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.383852333946774e-05,
+      "loss": 4.2197,
+      "step": 376320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.383013739195722e-05,
+      "loss": 4.2186,
+      "step": 376832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.1992,
+      "step": 377344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.2153,
+      "step": 377856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.209,
+      "step": 378368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2116,
+      "step": 378880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2171,
+      "step": 379392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2003,
+      "step": 379904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2183,
+      "step": 380416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.376309894828426e-05,
+      "loss": 4.2077,
+      "step": 380928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2199,
+      "step": 381440
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.2537760734558105,
+      "eval_runtime": 289.7497,
+      "eval_samples_per_second": 1316.968,
+      "eval_steps_per_second": 41.156,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.2022,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.213,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2147,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2081,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371278326322113e-05,
+      "loss": 4.2158,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370439731571061e-05,
+      "loss": 4.2013,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36960113682001e-05,
+      "loss": 4.2057,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368762542068958e-05,
+      "loss": 4.1948,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367923947317906e-05,
+      "loss": 4.2139,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367085352566854e-05,
+      "loss": 4.2115,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366246757815802e-05,
+      "loss": 4.208,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36540816306475e-05,
+      "loss": 4.2104,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364571206194071e-05,
+      "loss": 4.1999,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363732611443019e-05,
+      "loss": 4.1989,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362894016691967e-05,
+      "loss": 4.1981,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362055421940915e-05,
+      "loss": 4.1887,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361218465070236e-05,
+      "loss": 4.2043,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360379870319184e-05,
+      "loss": 4.1927,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359541275568132e-05,
+      "loss": 4.1962,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35870268081708e-05,
+      "loss": 4.2291,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3578640860660284e-05,
+      "loss": 4.1952,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357025491314976e-05,
+      "loss": 4.2038,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356186896563924e-05,
+      "loss": 4.2042,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355348301812872e-05,
+      "loss": 4.2025,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35450970706182e-05,
+      "loss": 4.1974,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3536711123107677e-05,
+      "loss": 4.1933,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528325175597157e-05,
+      "loss": 4.2036,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3519939228086636e-05,
+      "loss": 4.1836,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511569659379846e-05,
+      "loss": 4.1935,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503183711869326e-05,
+      "loss": 4.1895,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3494797764358805e-05,
+      "loss": 4.1934,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486411816848285e-05,
+      "loss": 4.2005,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34780422481415e-05,
+      "loss": 4.2018,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346965630063098e-05,
+      "loss": 4.1935,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346127035312046e-05,
+      "loss": 4.1997,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345288440560994e-05,
+      "loss": 4.2033,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344451483690315e-05,
+      "loss": 4.1892,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343612888939263e-05,
+      "loss": 4.202,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342774294188211e-05,
+      "loss": 4.1809,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341935699437159e-05,
+      "loss": 4.2041,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341097104686107e-05,
+      "loss": 4.1769,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340258509935055e-05,
+      "loss": 4.2085,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339419915184003e-05,
+      "loss": 4.195,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338581320432951e-05,
+      "loss": 4.1918,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337744363562272e-05,
+      "loss": 4.1964,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33690576881122e-05,
+      "loss": 4.1935,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360671740601686e-05,
+      "loss": 4.1949,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352285793091166e-05,
+      "loss": 4.1816,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343916224384375e-05,
+      "loss": 4.1756,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3335546655677584e-05,
+      "loss": 4.2039,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327160708167064e-05,
+      "loss": 4.1895,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3318774760656544e-05,
+      "loss": 4.1973,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3310388813146024e-05,
+      "loss": 4.1831,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330201924443923e-05,
+      "loss": 4.1894,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329363329692871e-05,
+      "loss": 4.1827,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328524734941819e-05,
+      "loss": 4.1918,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327686140190767e-05,
+      "loss": 4.1827,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326849183320089e-05,
+      "loss": 4.1986,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326010588569037e-05,
+      "loss": 4.1956,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325171993817985e-05,
+      "loss": 4.1795,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324333399066933e-05,
+      "loss": 4.1823,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323494804315881e-05,
+      "loss": 4.1933,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322656209564829e-05,
+      "loss": 4.1732,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321817614813776e-05,
+      "loss": 4.183,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320979020062724e-05,
+      "loss": 4.1878,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320140425311672e-05,
+      "loss": 4.1876,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319303468440994e-05,
+      "loss": 4.1693,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318464873689941e-05,
+      "loss": 4.1761,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317626278938889e-05,
+      "loss": 4.1697,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316787684187838e-05,
+      "loss": 4.1934,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315949089436786e-05,
+      "loss": 4.1849,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151121325661066e-05,
+      "loss": 4.1831,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3142735378150546e-05,
+      "loss": 4.1899,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134349430640026e-05,
+      "loss": 4.1783,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3125979861933235e-05,
+      "loss": 4.1992,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117593914422715e-05,
+      "loss": 4.1694,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109207966912195e-05,
+      "loss": 4.1852,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3100822019401675e-05,
+      "loss": 4.1829,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092436071891155e-05,
+      "loss": 4.1799,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084050124380635e-05,
+      "loss": 4.1856,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3075680555673844e-05,
+      "loss": 4.1779,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306729460816333e-05,
+      "loss": 4.1821,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305890866065281e-05,
+      "loss": 4.1886,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305052271314229e-05,
+      "loss": 4.169,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304213676563177e-05,
+      "loss": 4.1854,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303375081812125e-05,
+      "loss": 4.1747,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302536487061073e-05,
+      "loss": 4.1858,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301697892310021e-05,
+      "loss": 4.1739,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300859297558969e-05,
+      "loss": 4.1707,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.30002234068829e-05,
+      "loss": 4.1782,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299183745937238e-05,
+      "loss": 4.1776,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298346789066559e-05,
+      "loss": 4.1786,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297508194315507e-05,
+      "loss": 4.1655,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296669599564455e-05,
+      "loss": 4.1808,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295831004813403e-05,
+      "loss": 4.1874,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949924100623515e-05,
+      "loss": 4.1733,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2941538153112995e-05,
+      "loss": 4.1576,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2933152205602475e-05,
+      "loss": 4.1732,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292476625809195e-05,
+      "loss": 4.1756,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291638031058143e-05,
+      "loss": 4.1795,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290799436307091e-05,
+      "loss": 4.1802,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289960841556039e-05,
+      "loss": 4.169,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28912388468536e-05,
+      "loss": 4.166,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288285289934308e-05,
+      "loss": 4.1781,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287446695183256e-05,
+      "loss": 4.1553,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286608100432204e-05,
+      "loss": 4.1688,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285771143561525e-05,
+      "loss": 4.1673,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284932548810473e-05,
+      "loss": 4.1764,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284093954059421e-05,
+      "loss": 4.155,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283255359308369e-05,
+      "loss": 4.1666,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282416764557317e-05,
+      "loss": 4.1687,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281578169806265e-05,
+      "loss": 4.1658,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280739575055213e-05,
+      "loss": 4.1695,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279900980304161e-05,
+      "loss": 4.1585,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279064023433482e-05,
+      "loss": 4.1742,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27822542868243e-05,
+      "loss": 4.173,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277388471811751e-05,
+      "loss": 4.1629,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276549877060699e-05,
+      "loss": 4.1755,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275711282309647e-05,
+      "loss": 4.1595,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274872687558595e-05,
+      "loss": 4.1768,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274034092807543e-05,
+      "loss": 4.1613,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273197135936865e-05,
+      "loss": 4.1715,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2723585411858127e-05,
+      "loss": 4.1568,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715199464347607e-05,
+      "loss": 4.1666,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2706813516837087e-05,
+      "loss": 4.1629,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698427569326566e-05,
+      "loss": 4.1713,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690041621816046e-05,
+      "loss": 4.1594,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2681655674305526e-05,
+      "loss": 4.1673,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673269726795006e-05,
+      "loss": 4.1458,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664900158088215e-05,
+      "loss": 4.1812,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656514210577695e-05,
+      "loss": 4.1617,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648144641870904e-05,
+      "loss": 4.1681,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639758694360384e-05,
+      "loss": 4.1552,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263137274684987e-05,
+      "loss": 4.1725,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262298679933935e-05,
+      "loss": 4.1713,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261460085182883e-05,
+      "loss": 4.1609,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260621490431831e-05,
+      "loss": 4.1616,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597828956807784e-05,
+      "loss": 4.1683,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1683,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258107344059048e-05,
+      "loss": 4.1756,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257270387188369e-05,
+      "loss": 4.1466,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256431792437317e-05,
+      "loss": 4.1691,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255593197686265e-05,
+      "loss": 4.1566,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254754602935213e-05,
+      "loss": 4.1677,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253916008184161e-05,
+      "loss": 4.1641,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253077413433109e-05,
+      "loss": 4.1482,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522404565624305e-05,
+      "loss": 4.1717,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514018618113785e-05,
+      "loss": 4.1628,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250563267060326e-05,
+      "loss": 4.1656,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2123188972473145,
+      "eval_runtime": 288.1497,
+      "eval_samples_per_second": 1324.28,
+      "eval_steps_per_second": 41.385,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249724672309274e-05,
+      "loss": 4.1574,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248886077558222e-05,
+      "loss": 4.158,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24804748280717e-05,
+      "loss": 4.1675,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247208888056118e-05,
+      "loss": 4.1612,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246370293305066e-05,
+      "loss": 4.165,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245531698554014e-05,
+      "loss": 4.1579,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244694741683335e-05,
+      "loss": 4.1512,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243857784812656e-05,
+      "loss": 4.1523,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243019190061604e-05,
+      "loss": 4.1597,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242180595310552e-05,
+      "loss": 4.1648,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413420005595e-05,
+      "loss": 4.1584,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240503405808448e-05,
+      "loss": 4.1617,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239666448937769e-05,
+      "loss": 4.1494,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238827854186717e-05,
+      "loss": 4.1552,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237989259435665e-05,
+      "loss": 4.1519,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237150664684613e-05,
+      "loss": 4.1403,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236312069933561e-05,
+      "loss": 4.1577,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235473475182509e-05,
+      "loss": 4.1416,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234634880431457e-05,
+      "loss": 4.1479,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233796285680405e-05,
+      "loss": 4.1816,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232957690929353e-05,
+      "loss": 4.1498,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232119096178301e-05,
+      "loss": 4.1555,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.155,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2304451824369436e-05,
+      "loss": 4.1617,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1418,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2287679929348396e-05,
+      "loss": 4.1485,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2279293981837876e-05,
+      "loss": 4.1526,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2270924413131085e-05,
+      "loss": 4.1419,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2262538465620565e-05,
+      "loss": 4.1475,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2254152518110045e-05,
+      "loss": 4.1418,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2245766570599525e-05,
+      "loss": 4.1498,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2237380623089005e-05,
+      "loss": 4.1554,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2228994675578485e-05,
+      "loss": 4.1514,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2220608728067965e-05,
+      "loss": 4.1466,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212222780557445e-05,
+      "loss": 4.1507,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2203836833046925e-05,
+      "loss": 4.1565,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219548364314387e-05,
+      "loss": 4.143,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218709769563335e-05,
+      "loss": 4.1605,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217871174812283e-05,
+      "loss": 4.1348,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217032580061231e-05,
+      "loss": 4.154,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216193985310179e-05,
+      "loss": 4.1318,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215355390559126e-05,
+      "loss": 4.1632,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214518433688448e-05,
+      "loss": 4.1465,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213679838937396e-05,
+      "loss": 4.1504,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212841244186344e-05,
+      "loss": 4.1482,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212002649435292e-05,
+      "loss": 4.1472,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21116405468424e-05,
+      "loss": 4.1511,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210325459933188e-05,
+      "loss": 4.1394,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209486865182136e-05,
+      "loss": 4.1269,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208648270431084e-05,
+      "loss": 4.1594,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207811313560405e-05,
+      "loss": 4.1415,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206972718809353e-05,
+      "loss": 4.1561,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206134124058301e-05,
+      "loss": 4.1341,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2052971671876216e-05,
+      "loss": 4.1425,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2044585724365696e-05,
+      "loss": 4.1335,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1493,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.1408,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.1513,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1503,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1343,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199427003930258e-05,
+      "loss": 4.1338,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1536,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197751452308527e-05,
+      "loss": 4.1282,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196912857557475e-05,
+      "loss": 4.1372,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196074262806423e-05,
+      "loss": 4.1416,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195237305935744e-05,
+      "loss": 4.148,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194398711184692e-05,
+      "loss": 4.1255,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19356011643364e-05,
+      "loss": 4.1301,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192721521682588e-05,
+      "loss": 4.1287,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1918862026922826e-05,
+      "loss": 4.1398,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1910476079412306e-05,
+      "loss": 4.1448,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.1399,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.142,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1331,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1876932289370226e-05,
+      "loss": 4.1583,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1868562720663435e-05,
+      "loss": 4.1259,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860176773152915e-05,
+      "loss": 4.1413,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1851790825642395e-05,
+      "loss": 4.1411,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1843404878131875e-05,
+      "loss": 4.136,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835018930621355e-05,
+      "loss": 4.1434,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1826632983110834e-05,
+      "loss": 4.137,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818247035600314e-05,
+      "loss": 4.1336,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1809861088089794e-05,
+      "loss": 4.1454,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180149151938301e-05,
+      "loss": 4.131,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179312195067622e-05,
+      "loss": 4.1378,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.17847360031657e-05,
+      "loss": 4.1339,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177635005565518e-05,
+      "loss": 4.1428,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1304,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1266,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175120859192735e-05,
+      "loss": 4.1337,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174282264441683e-05,
+      "loss": 4.1359,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173443669690631e-05,
+      "loss": 4.1371,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172605074939579e-05,
+      "loss": 4.1231,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171766480188527e-05,
+      "loss": 4.1392,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170927885437475e-05,
+      "loss": 4.1424,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170089290686423e-05,
+      "loss": 4.1345,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169250695935371e-05,
+      "loss": 4.1164,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168412101184319e-05,
+      "loss": 4.1303,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16757514431364e-05,
+      "loss": 4.1289,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166736549562588e-05,
+      "loss": 4.1413,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165897954811536e-05,
+      "loss": 4.1386,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165059360060484e-05,
+      "loss": 4.1262,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164220765309432e-05,
+      "loss": 4.1252,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1363,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162546851568074e-05,
+      "loss": 4.1155,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161708256817022e-05,
+      "loss": 4.1309,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16086966206597e-05,
+      "loss": 4.1212,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160031067314918e-05,
+      "loss": 4.1392,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159192472563866e-05,
+      "loss": 4.1127,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1235,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157515283061762e-05,
+      "loss": 4.1238,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.1295,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155839731440031e-05,
+      "loss": 4.1247,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155001136688979e-05,
+      "loss": 4.1186,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154162541937927e-05,
+      "loss": 4.1292,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153323947186875e-05,
+      "loss": 4.1314,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1524886281965695e-05,
+      "loss": 4.1211,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151650033445517e-05,
+      "loss": 4.1383,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1508114386944655e-05,
+      "loss": 4.1185,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1499728439434135e-05,
+      "loss": 4.1312,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491342491923615e-05,
+      "loss": 4.1227,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1482956544413095e-05,
+      "loss": 4.1289,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1474570596902575e-05,
+      "loss": 4.1183,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.1249,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1457798701881535e-05,
+      "loss": 4.1287,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1449429133174744e-05,
+      "loss": 4.1244,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1441043185664224e-05,
+      "loss": 4.1177,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1277,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142428766944691e-05,
+      "loss": 4.1092,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141590172193639e-05,
+      "loss": 4.1384,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140751577442587e-05,
+      "loss": 4.1235,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139912982691535e-05,
+      "loss": 4.1266,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139074387940483e-05,
+      "loss": 4.118,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138235793189432e-05,
+      "loss": 4.1355,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13739719843838e-05,
+      "loss": 4.1261,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.121,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1212,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1268,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134044457314545e-05,
+      "loss": 4.1258,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133205862563493e-05,
+      "loss": 4.1364,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132367267812441e-05,
+      "loss": 4.1058,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131528673061388e-05,
+      "loss": 4.1277,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1202,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129853121439658e-05,
+      "loss": 4.1288,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129014526688606e-05,
+      "loss": 4.1202,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128175931937554e-05,
+      "loss": 4.1136,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127337337186502e-05,
+      "loss": 4.1288,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126500380315823e-05,
+      "loss": 4.1233,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1256617855647707e-05,
+      "loss": 4.1242,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.181349754333496,
+      "eval_runtime": 287.3473,
+      "eval_samples_per_second": 1327.978,
+      "eval_steps_per_second": 41.5,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248231908137186e-05,
+      "loss": 4.1217,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1239845960626666e-05,
+      "loss": 4.1183,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1231460013116146e-05,
+      "loss": 4.1269,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1223074065605626e-05,
+      "loss": 4.1253,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1214688118095106e-05,
+      "loss": 4.1278,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206302170584586e-05,
+      "loss": 4.1156,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1197916223074066e-05,
+      "loss": 4.1121,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1189530275563546e-05,
+      "loss": 4.1096,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1181144328053026e-05,
+      "loss": 4.1245,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1172758380542506e-05,
+      "loss": 4.1309,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1164372433031986e-05,
+      "loss": 4.1165,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115598648552147e-05,
+      "loss": 4.1219,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114760053801095e-05,
+      "loss": 4.1131,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1139214590500426e-05,
+      "loss": 4.1164,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1130828642989906e-05,
+      "loss": 4.1114,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122442695479386e-05,
+      "loss": 4.1069,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1114056747968866e-05,
+      "loss": 4.1152,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1105670800458346e-05,
+      "loss": 4.1087,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1097301231751555e-05,
+      "loss": 4.1074,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1088915284241035e-05,
+      "loss": 4.1406,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1080529336730515e-05,
+      "loss": 4.1164,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072143389219995e-05,
+      "loss": 4.1146,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106377382051321e-05,
+      "loss": 4.1168,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105538787300269e-05,
+      "loss": 4.1202,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104700192549217e-05,
+      "loss": 4.1041,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103861597798165e-05,
+      "loss": 4.1128,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103023003047113e-05,
+      "loss": 4.1143,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102184408296061e-05,
+      "loss": 4.1074,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101345813545009e-05,
+      "loss": 4.1042,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100507218793957e-05,
+      "loss": 4.1058,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099668624042905e-05,
+      "loss": 4.1108,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098833305052599e-05,
+      "loss": 4.1192,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097994710301547e-05,
+      "loss": 4.1146,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097156115550495e-05,
+      "loss": 4.1058,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096317520799443e-05,
+      "loss": 4.1155,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095478926048391e-05,
+      "loss": 4.1156,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946419691777124e-05,
+      "loss": 4.1108,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938033744266604e-05,
+      "loss": 4.119,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0929647796756084e-05,
+      "loss": 4.1,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921261849245564e-05,
+      "loss": 4.1162,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0912875901735044e-05,
+      "loss": 4.0995,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904489954224524e-05,
+      "loss": 4.1241,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896104006714004e-05,
+      "loss": 4.1123,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088775081681094e-05,
+      "loss": 4.1126,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087936486930042e-05,
+      "loss": 4.1144,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08709789217899e-05,
+      "loss": 4.1066,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086259297427938e-05,
+      "loss": 4.1154,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085420702676886e-05,
+      "loss": 4.1048,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084582107925834e-05,
+      "loss": 4.0851,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083743513174783e-05,
+      "loss": 4.1238,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082904918423731e-05,
+      "loss": 4.1043,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082066323672678e-05,
+      "loss": 4.1198,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081227728921626e-05,
+      "loss": 4.1022,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080389134170574e-05,
+      "loss": 4.1059,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079550539419522e-05,
+      "loss": 4.0958,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078715220429217e-05,
+      "loss": 4.1139,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0778766256781647e-05,
+      "loss": 4.1071,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770396688074856e-05,
+      "loss": 4.1121,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0762010740564336e-05,
+      "loss": 4.1197,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0753624793053816e-05,
+      "loss": 4.0984,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745238845543296e-05,
+      "loss": 4.0988,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073685289803278e-05,
+      "loss": 4.1177,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0728466950522255e-05,
+      "loss": 4.0943,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720081003011735e-05,
+      "loss": 4.1001,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0711695055501215e-05,
+      "loss": 4.1059,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703309107990695e-05,
+      "loss": 4.1119,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0694923160480175e-05,
+      "loss": 4.0911,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0686537212969655e-05,
+      "loss": 4.0961,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0678151265459135e-05,
+      "loss": 4.0903,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0669765317948615e-05,
+      "loss": 4.1063,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0661395749241824e-05,
+      "loss": 4.1113,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653009801731304e-05,
+      "loss": 4.1055,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0644623854220784e-05,
+      "loss": 4.1025,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636237906710264e-05,
+      "loss": 4.1024,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062785195919975e-05,
+      "loss": 4.1217,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061948239049296e-05,
+      "loss": 4.0918,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061109644298244e-05,
+      "loss": 4.1107,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060271049547192e-05,
+      "loss": 4.1053,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05943245479614e-05,
+      "loss": 4.0957,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058595497925461e-05,
+      "loss": 4.1114,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057758541054782e-05,
+      "loss": 4.102,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05691994630373e-05,
+      "loss": 4.1002,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056081351552678e-05,
+      "loss": 4.1105,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055242756801626e-05,
+      "loss": 4.0923,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054404162050574e-05,
+      "loss": 4.1009,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053565567299522e-05,
+      "loss": 4.1005,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527286104288434e-05,
+      "loss": 4.1079,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0518900156777914e-05,
+      "loss": 4.0953,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510514209267394e-05,
+      "loss": 4.0964,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502128261756874e-05,
+      "loss": 4.0927,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0493742314246353e-05,
+      "loss": 4.1036,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048537274553956e-05,
+      "loss": 4.1053,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047698679802904e-05,
+      "loss": 4.0882,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046860085051852e-05,
+      "loss": 4.1065,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460214903008e-05,
+      "loss": 4.1061,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045182895549748e-05,
+      "loss": 4.0986,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044344300798696e-05,
+      "loss": 4.083,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043505706047644e-05,
+      "loss": 4.096,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042668749176965e-05,
+      "loss": 4.0929,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041830154425914e-05,
+      "loss": 4.1086,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040991559674862e-05,
+      "loss": 4.104,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040152964923809e-05,
+      "loss": 4.0926,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039314370172757e-05,
+      "loss": 4.0943,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038477413302079e-05,
+      "loss": 4.0953,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0376404564313996e-05,
+      "loss": 4.0886,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0368018616803476e-05,
+      "loss": 4.0954,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0359632669292956e-05,
+      "loss": 4.087,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351246721782436e-05,
+      "loss": 4.1068,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034286077427191e-05,
+      "loss": 4.0785,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0334491205565125e-05,
+      "loss": 4.0904,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0326105258054605e-05,
+      "loss": 4.0883,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031771931054409e-05,
+      "loss": 4.0983,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0309333363033565e-05,
+      "loss": 4.0872,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0300947415523045e-05,
+      "loss": 4.0876,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0292561468012525e-05,
+      "loss": 4.0974,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284175520502005e-05,
+      "loss": 4.0995,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0275789572991485e-05,
+      "loss": 4.0848,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267420004284694e-05,
+      "loss": 4.1003,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259034056774174e-05,
+      "loss": 4.0877,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250648109263654e-05,
+      "loss": 4.0962,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242262161753134e-05,
+      "loss": 4.0918,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233876214242614e-05,
+      "loss": 4.0932,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225490266732094e-05,
+      "loss": 4.0848,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217104319221574e-05,
+      "loss": 4.0938,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020871837171106e-05,
+      "loss": 4.097,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020034880300427e-05,
+      "loss": 4.0924,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019196285549375e-05,
+      "loss": 4.0865,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018357690798323e-05,
+      "loss": 4.0915,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017519096047271e-05,
+      "loss": 4.076,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016683777056965e-05,
+      "loss": 4.1078,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015845182305913e-05,
+      "loss": 4.0864,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015006587554861e-05,
+      "loss": 4.0938,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014167992803809e-05,
+      "loss": 4.0832,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013329398052757e-05,
+      "loss": 4.1017,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012490803301705e-05,
+      "loss": 4.0965,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.0891,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010815251679974e-05,
+      "loss": 4.0907,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009976656928922e-05,
+      "loss": 4.0901,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00913806217787e-05,
+      "loss": 4.0913,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008299467426818e-05,
+      "loss": 4.1087,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007460872675766e-05,
+      "loss": 4.0702,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006622277924714e-05,
+      "loss": 4.099,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005783683173662e-05,
+      "loss": 4.0861,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004946726302983e-05,
+      "loss": 4.0945,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004108131551931e-05,
+      "loss": 4.0847,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003269536800879e-05,
+      "loss": 4.0853,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002430942049827e-05,
+      "loss": 4.0943,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001593985179148e-05,
+      "loss": 4.0933,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000755390428097e-05,
+      "loss": 4.0913,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.155541896820068,
+      "eval_runtime": 293.969,
+      "eval_samples_per_second": 1298.065,
+      "eval_steps_per_second": 40.565,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999916795677045e-05,
+      "loss": 4.0901,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999078200925993e-05,
+      "loss": 4.0865,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99823960617494e-05,
+      "loss": 4.094,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997402649304262e-05,
+      "loss": 4.0916,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99656405455321e-05,
+      "loss": 4.0985,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9957270976825306e-05,
+      "loss": 4.0839,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9948885029314786e-05,
+      "loss": 4.0832,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9940499081804266e-05,
+      "loss": 4.0814,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9932113134293746e-05,
+      "loss": 4.088,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992372718678322e-05,
+      "loss": 4.096,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9915341239272706e-05,
+      "loss": 4.0843,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9906955291762185e-05,
+      "loss": 4.0894,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989860210185913e-05,
+      "loss": 4.0865,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989021615434861e-05,
+      "loss": 4.0804,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988183020683809e-05,
+      "loss": 4.0827,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987344425932757e-05,
+      "loss": 4.0729,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9865058311817044e-05,
+      "loss": 4.0843,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9856672364306523e-05,
+      "loss": 4.0747,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9848286416796003e-05,
+      "loss": 4.0793,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9839900469285483e-05,
+      "loss": 4.1029,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983151452177496e-05,
+      "loss": 4.0909,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982312857426444e-05,
+      "loss": 4.0833,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981474262675392e-05,
+      "loss": 4.0857,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98063566792434e-05,
+      "loss": 4.0911,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979798711053662e-05,
+      "loss": 4.0693,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97896011630261e-05,
+      "loss": 4.0839,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978121521551558e-05,
+      "loss": 4.0859,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977282926800506e-05,
+      "loss": 4.0721,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976444332049454e-05,
+      "loss": 4.0722,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.975605737298402e-05,
+      "loss": 4.0755,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974768780427723e-05,
+      "loss": 4.0801,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973930185676671e-05,
+      "loss": 4.0895,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973091590925619e-05,
+      "loss": 4.0826,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.972252996174567e-05,
+      "loss": 4.079,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9714176771842606e-05,
+      "loss": 4.0845,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970579082433209e-05,
+      "loss": 4.0843,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969740487682157e-05,
+      "loss": 4.0811,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968901892931105e-05,
+      "loss": 4.0845,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968063298180053e-05,
+      "loss": 4.0684,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967224703429001e-05,
+      "loss": 4.0816,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966386108677949e-05,
+      "loss": 4.0701,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96554915180727e-05,
+      "loss": 4.0916,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964710557056218e-05,
+      "loss": 4.0835,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963871962305166e-05,
+      "loss": 4.078,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963033367554114e-05,
+      "loss": 4.0858,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962194772803062e-05,
+      "loss": 4.0777,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96135617805201e-05,
+      "loss": 4.0879,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960517583300958e-05,
+      "loss": 4.0764,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959678988549906e-05,
+      "loss": 4.0499,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958842031679228e-05,
+      "loss": 4.0917,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958003436928176e-05,
+      "loss": 4.0761,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957164842177124e-05,
+      "loss": 4.0934,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956326247426071e-05,
+      "loss": 4.0712,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955487652675019e-05,
+      "loss": 4.0731,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9546506958043406e-05,
+      "loss": 4.0675,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953812101053288e-05,
+      "loss": 4.0814,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952973506302236e-05,
+      "loss": 4.0818,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952134911551184e-05,
+      "loss": 4.0801,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951296316800132e-05,
+      "loss": 4.0923,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950459359929453e-05,
+      "loss": 4.0716,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9496207651784015e-05,
+      "loss": 4.0646,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9487821704273495e-05,
+      "loss": 4.0895,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9479435756762975e-05,
+      "loss": 4.0643,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9471049809252455e-05,
+      "loss": 4.0688,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94626966193494e-05,
+      "loss": 4.0747,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945431067183888e-05,
+      "loss": 4.0851,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944592472432835e-05,
+      "loss": 4.0626,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943753877681783e-05,
+      "loss": 4.0689,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942916920811105e-05,
+      "loss": 4.0612,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942078326060053e-05,
+      "loss": 4.0742,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941239731309e-05,
+      "loss": 4.0824,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940401136557948e-05,
+      "loss": 4.0731,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939562541806897e-05,
+      "loss": 4.0759,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938723947055845e-05,
+      "loss": 4.072,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937885352304793e-05,
+      "loss": 4.0957,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937046757553741e-05,
+      "loss": 4.0631,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936208162802689e-05,
+      "loss": 4.0777,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935369568051637e-05,
+      "loss": 4.0768,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.934530973300585e-05,
+      "loss": 4.069,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933692378549533e-05,
+      "loss": 4.0782,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932857059559227e-05,
+      "loss": 4.074,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9320201026885476e-05,
+      "loss": 4.0718,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9311815079374956e-05,
+      "loss": 4.0774,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9303429131864436e-05,
+      "loss": 4.0687,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929504318435392e-05,
+      "loss": 4.0683,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92866572368434e-05,
+      "loss": 4.0735,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927827128933288e-05,
+      "loss": 4.0782,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926988534182236e-05,
+      "loss": 4.0658,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926149939431184e-05,
+      "loss": 4.0684,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925311344680132e-05,
+      "loss": 4.0644,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92447274992908e-05,
+      "loss": 4.079,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923634155178028e-05,
+      "loss": 4.0738,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922795560426976e-05,
+      "loss": 4.06,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921956965675924e-05,
+      "loss": 4.0765,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921120008805245e-05,
+      "loss": 4.0769,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920281414054193e-05,
+      "loss": 4.0715,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919442819303141e-05,
+      "loss": 4.0541,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918604224552089e-05,
+      "loss": 4.0666,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91776726768141e-05,
+      "loss": 4.0635,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9169303108107316e-05,
+      "loss": 4.0786,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9160917160596796e-05,
+      "loss": 4.0751,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9152531213086276e-05,
+      "loss": 4.0702,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9144145265575756e-05,
+      "loss": 4.0662,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9135775696868965e-05,
+      "loss": 4.0648,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9127389749358445e-05,
+      "loss": 4.0623,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9119003801847925e-05,
+      "loss": 4.064,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9110617854337405e-05,
+      "loss": 4.0599,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9102231906826885e-05,
+      "loss": 4.0756,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9093845959316365e-05,
+      "loss": 4.0544,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9085476390609574e-05,
+      "loss": 4.0563,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9077090443099054e-05,
+      "loss": 4.0628,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906870449558854e-05,
+      "loss": 4.0633,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9060318548078014e-05,
+      "loss": 4.0633,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9051932600567494e-05,
+      "loss": 4.0574,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9043546653056974e-05,
+      "loss": 4.0676,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9035160705546453e-05,
+      "loss": 4.0743,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9026774758035933e-05,
+      "loss": 4.0562,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901838881052541e-05,
+      "loss": 4.0702,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901001924181862e-05,
+      "loss": 4.0617,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90016332943081e-05,
+      "loss": 4.0691,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899324734679758e-05,
+      "loss": 4.0662,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898486139928706e-05,
+      "loss": 4.0616,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897649183058028e-05,
+      "loss": 4.0594,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896810588306976e-05,
+      "loss": 4.0653,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895971993555924e-05,
+      "loss": 4.0679,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895133398804872e-05,
+      "loss": 4.0622,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.894296441934193e-05,
+      "loss": 4.0592,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8934594850635136e-05,
+      "loss": 4.0636,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8926208903124616e-05,
+      "loss": 4.0483,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917822955614096e-05,
+      "loss": 4.0787,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909437008103576e-05,
+      "loss": 4.0567,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8901067439396785e-05,
+      "loss": 4.0703,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8892681491886265e-05,
+      "loss": 4.0526,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8884295544375745e-05,
+      "loss": 4.0749,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887590959686523e-05,
+      "loss": 4.0681,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886752364935471e-05,
+      "loss": 4.0608,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885913770184419e-05,
+      "loss": 4.0624,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885075175433367e-05,
+      "loss": 4.0648,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884236580682315e-05,
+      "loss": 4.066,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883397985931263e-05,
+      "loss": 4.0772,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882561029060584e-05,
+      "loss": 4.0472,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881722434309532e-05,
+      "loss": 4.0723,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88088383955848e-05,
+      "loss": 4.0564,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880045244807428e-05,
+      "loss": 4.068,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879208287936749e-05,
+      "loss": 4.0609,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878369693185697e-05,
+      "loss": 4.0579,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877531098434645e-05,
+      "loss": 4.066,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876692503683593e-05,
+      "loss": 4.0668,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8758539089325416e-05,
+      "loss": 4.0629,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.136168956756592,
+      "eval_runtime": 294.2435,
+      "eval_samples_per_second": 1296.854,
+      "eval_steps_per_second": 40.528,
+      "step": 686880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8750153141814896e-05,
+      "loss": 4.0596,
+      "step": 687104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8741767194304376e-05,
+      "loss": 4.0597,
+      "step": 687616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.873338124679385e-05,
+      "loss": 4.0677,
+      "step": 688128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8725011678087065e-05,
+      "loss": 4.0627,
+      "step": 688640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8716625730576545e-05,
+      "loss": 4.0698,
+      "step": 689152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8708239783066025e-05,
+      "loss": 4.0571,
+      "step": 689664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.86998538355555e-05,
+      "loss": 4.0576,
+      "step": 690176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8691484266848714e-05,
+      "loss": 4.0562,
+      "step": 690688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.868311469814192e-05,
+      "loss": 4.0603,
+      "step": 691200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.86747287506314e-05,
+      "loss": 4.0688,
+      "step": 691712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.866634280312088e-05,
+      "loss": 4.0596,
+      "step": 692224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.86579732344141e-05,
+      "loss": 4.0613,
+      "step": 692736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.864960366570731e-05,
+      "loss": 4.0629,
+      "step": 693248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.864121771819679e-05,
+      "loss": 4.052,
+      "step": 693760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.863283177068627e-05,
+      "loss": 4.056,
+      "step": 694272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.862444582317575e-05,
+      "loss": 4.0471,
+      "step": 694784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.861605987566523e-05,
+      "loss": 4.0528,
+      "step": 695296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.860767392815471e-05,
+      "loss": 4.0499,
+      "step": 695808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.859928798064419e-05,
+      "loss": 4.0515,
+      "step": 696320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.859090203313367e-05,
+      "loss": 4.0722,
+      "step": 696832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.858251608562314e-05,
+      "loss": 4.0666,
+      "step": 697344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.857413013811262e-05,
+      "loss": 4.0598,
+      "step": 697856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.856574419060211e-05,
+      "loss": 4.0589,
+      "step": 698368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.855735824309159e-05,
+      "loss": 4.0648,
+      "step": 698880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.854897229558107e-05,
+      "loss": 4.0427,
+      "step": 699392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.854058634807055e-05,
+      "loss": 4.0573,
+      "step": 699904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.853220040056003e-05,
+      "loss": 4.0597,
+      "step": 700416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.852381445304951e-05,
+      "loss": 4.0438,
+      "step": 700928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.851544488434272e-05,
+      "loss": 4.0517,
+      "step": 701440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8507075315635926e-05,
+      "loss": 4.0482,
+      "step": 701952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8498689368125406e-05,
+      "loss": 4.0566,
+      "step": 702464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8490319799418615e-05,
+      "loss": 4.059,
+      "step": 702976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8481933851908095e-05,
+      "loss": 4.0548,
+      "step": 703488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8473547904397575e-05,
+      "loss": 4.0611,
+      "step": 704000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8465161956887055e-05,
+      "loss": 4.0521,
+      "step": 704512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.845677600937654e-05,
+      "loss": 4.0577,
+      "step": 705024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.844840644066975e-05,
+      "loss": 4.0622,
+      "step": 705536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.844002049315923e-05,
+      "loss": 4.0524,
+      "step": 706048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.843163454564871e-05,
+      "loss": 4.0435,
+      "step": 706560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.842324859813819e-05,
+      "loss": 4.057,
+      "step": 707072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.841486265062767e-05,
+      "loss": 4.0434,
+      "step": 707584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.840647670311715e-05,
+      "loss": 4.0652,
+      "step": 708096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.839810713441036e-05,
+      "loss": 4.0585,
+      "step": 708608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.838972118689984e-05,
+      "loss": 4.0505,
+      "step": 709120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.838133523938932e-05,
+      "loss": 4.061,
+      "step": 709632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.83729492918788e-05,
+      "loss": 4.0531,
+      "step": 710144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.836456334436828e-05,
+      "loss": 4.0594,
+      "step": 710656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.835617739685776e-05,
+      "loss": 4.0565,
+      "step": 711168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.834779144934724e-05,
+      "loss": 4.0211,
+      "step": 711680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8339405501836726e-05,
+      "loss": 4.0661,
+      "step": 712192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8331019554326206e-05,
+      "loss": 4.0522,
+      "step": 712704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8322649985619415e-05,
+      "loss": 4.061,
+      "step": 713216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8314264038108895e-05,
+      "loss": 4.049,
+      "step": 713728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8305878090598375e-05,
+      "loss": 4.0499,
+      "step": 714240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8297492143087855e-05,
+      "loss": 4.0404,
+      "step": 714752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8289122574381064e-05,
+      "loss": 4.0558,
+      "step": 715264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8280736626870544e-05,
+      "loss": 4.054,
+      "step": 715776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8272350679360024e-05,
+      "loss": 4.0533,
+      "step": 716288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8263964731849504e-05,
+      "loss": 4.0698,
+      "step": 716800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.825557878433898e-05,
+      "loss": 4.0474,
+      "step": 717312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8247192836828464e-05,
+      "loss": 4.035,
+      "step": 717824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.823882326812168e-05,
+      "loss": 4.0656,
+      "step": 718336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.823043732061116e-05,
+      "loss": 4.0397,
+      "step": 718848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.822205137310063e-05,
+      "loss": 4.0397,
+      "step": 719360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.821366542559011e-05,
+      "loss": 4.0557,
+      "step": 719872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.820527947807959e-05,
+      "loss": 4.055,
+      "step": 720384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.819689353056907e-05,
+      "loss": 4.0391,
+      "step": 720896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.818850758305855e-05,
+      "loss": 4.0425,
+      "step": 721408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.818013801435176e-05,
+      "loss": 4.0359,
+      "step": 721920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.817175206684124e-05,
+      "loss": 4.0465,
+      "step": 722432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.816336611933072e-05,
+      "loss": 4.059,
+      "step": 722944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.81549801718202e-05,
+      "loss": 4.0475,
+      "step": 723456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.814659422430968e-05,
+      "loss": 4.0502,
+      "step": 723968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.813820827679916e-05,
+      "loss": 4.0437,
+      "step": 724480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.812982232928865e-05,
+      "loss": 4.069,
+      "step": 724992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.812145276058186e-05,
+      "loss": 4.0422,
+      "step": 725504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.811306681307134e-05,
+      "loss": 4.0545,
+      "step": 726016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.810468086556082e-05,
+      "loss": 4.0497,
+      "step": 726528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.80962949180503e-05,
+      "loss": 4.0453,
+      "step": 727040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.808790897053978e-05,
+      "loss": 4.051,
+      "step": 727552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.807952302302926e-05,
+      "loss": 4.0534,
+      "step": 728064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.807113707551874e-05,
+      "loss": 4.0468,
+      "step": 728576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.806275112800822e-05,
+      "loss": 4.0547,
+      "step": 729088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8054381559301426e-05,
+      "loss": 4.0445,
+      "step": 729600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8045995611790906e-05,
+      "loss": 4.0432,
+      "step": 730112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.8037609664280386e-05,
+      "loss": 4.0468,
+      "step": 730624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.80292400955736e-05,
+      "loss": 4.0514,
+      "step": 731136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.802085414806308e-05,
+      "loss": 4.0418,
+      "step": 731648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.801248457935629e-05,
+      "loss": 4.046,
+      "step": 732160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.800409863184577e-05,
+      "loss": 4.042,
+      "step": 732672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.799571268433525e-05,
+      "loss": 4.0499,
+      "step": 733184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.798732673682473e-05,
+      "loss": 4.0497,
+      "step": 733696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.797894078931421e-05,
+      "loss": 4.0347,
+      "step": 734208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.797055484180369e-05,
+      "loss": 4.0533,
+      "step": 734720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7962168894293164e-05,
+      "loss": 4.048,
+      "step": 735232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7953782946782644e-05,
+      "loss": 4.0502,
+      "step": 735744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.794541337807586e-05,
+      "loss": 4.0297,
+      "step": 736256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.793702743056534e-05,
+      "loss": 4.0409,
+      "step": 736768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.792864148305482e-05,
+      "loss": 4.0411,
+      "step": 737280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.79202555355443e-05,
+      "loss": 4.0554,
+      "step": 737792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.791186958803378e-05,
+      "loss": 4.0491,
+      "step": 738304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.790348364052326e-05,
+      "loss": 4.0472,
+      "step": 738816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.789509769301274e-05,
+      "loss": 4.0439,
+      "step": 739328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.788671174550222e-05,
+      "loss": 4.0395,
+      "step": 739840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7878358555599164e-05,
+      "loss": 4.0353,
+      "step": 740352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.786997260808864e-05,
+      "loss": 4.0438,
+      "step": 740864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.786158666057812e-05,
+      "loss": 4.0358,
+      "step": 741376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.78532007130676e-05,
+      "loss": 4.0471,
+      "step": 741888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.784481476555708e-05,
+      "loss": 4.0331,
+      "step": 742400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.783642881804656e-05,
+      "loss": 4.0309,
+      "step": 742912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.782804287053604e-05,
+      "loss": 4.0372,
+      "step": 743424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.781965692302552e-05,
+      "loss": 4.042,
+      "step": 743936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7811270975515004e-05,
+      "loss": 4.0381,
+      "step": 744448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.780290140680821e-05,
+      "loss": 4.0332,
+      "step": 744960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.779451545929769e-05,
+      "loss": 4.0474,
+      "step": 745472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.77861458905909e-05,
+      "loss": 4.049,
+      "step": 745984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.777775994308038e-05,
+      "loss": 4.0357,
+      "step": 746496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.776937399556986e-05,
+      "loss": 4.0426,
+      "step": 747008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.776098804805934e-05,
+      "loss": 4.042,
+      "step": 747520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.775260210054882e-05,
+      "loss": 4.0402,
+      "step": 748032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.77442161530383e-05,
+      "loss": 4.0443,
+      "step": 748544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.773583020552778e-05,
+      "loss": 4.0361,
+      "step": 749056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.772744425801726e-05,
+      "loss": 4.0392,
+      "step": 749568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.771907468931047e-05,
+      "loss": 4.0376,
+      "step": 750080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.771068874179996e-05,
+      "loss": 4.0447,
+      "step": 750592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.770230279428944e-05,
+      "loss": 4.0413,
+      "step": 751104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.769391684677892e-05,
+      "loss": 4.0328,
+      "step": 751616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7685563656875856e-05,
+      "loss": 4.0399,
+      "step": 752128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7677177709365336e-05,
+      "loss": 4.0254,
+      "step": 752640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7668791761854816e-05,
+      "loss": 4.0539,
+      "step": 753152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7660405814344296e-05,
+      "loss": 4.0343,
+      "step": 753664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7652019866833776e-05,
+      "loss": 4.0488,
+      "step": 754176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7643633919323256e-05,
+      "loss": 4.0312,
+      "step": 754688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7635247971812735e-05,
+      "loss": 4.0507,
+      "step": 755200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7626862024302215e-05,
+      "loss": 4.0462,
+      "step": 755712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7618492455595425e-05,
+      "loss": 4.0366,
+      "step": 756224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.761012288688864e-05,
+      "loss": 4.0391,
+      "step": 756736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.760173693937812e-05,
+      "loss": 4.0381,
+      "step": 757248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.75933509918676e-05,
+      "loss": 4.0479,
+      "step": 757760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.758498142316081e-05,
+      "loss": 4.051,
+      "step": 758272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.757659547565029e-05,
+      "loss": 4.0238,
+      "step": 758784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.756820952813977e-05,
+      "loss": 4.0434,
+      "step": 759296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.755982358062925e-05,
+      "loss": 4.0373,
+      "step": 759808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.755143763311873e-05,
+      "loss": 4.045,
+      "step": 760320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.754305168560821e-05,
+      "loss": 4.0357,
+      "step": 760832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.753466573809769e-05,
+      "loss": 4.0396,
+      "step": 761344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.75262961693909e-05,
+      "loss": 4.0396,
+      "step": 761856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.751791022188038e-05,
+      "loss": 4.0429,
+      "step": 762368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.7509524274369865e-05,
+      "loss": 4.0421,
+      "step": 762880
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.120429515838623,
+      "eval_runtime": 294.5077,
+      "eval_samples_per_second": 1295.691,
+      "eval_steps_per_second": 40.491,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501138326859345e-05,
+      "loss": 4.0333,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7492752379348825e-05,
+      "loss": 4.0356,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74843664318383e-05,
+      "loss": 4.0438,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747598048432778e-05,
+      "loss": 4.0417,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746759453681726e-05,
+      "loss": 4.0455,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745920858930674e-05,
+      "loss": 4.0356,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745082264179622e-05,
+      "loss": 4.035,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74424366942857e-05,
+      "loss": 4.0315,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743405074677518e-05,
+      "loss": 4.0354,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742566479926466e-05,
+      "loss": 4.0486,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741727885175414e-05,
+      "loss": 4.033,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740889290424362e-05,
+      "loss": 4.0356,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740052333553683e-05,
+      "loss": 4.0446,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392137388026313e-05,
+      "loss": 4.029,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7383751440515793e-05,
+      "loss": 4.036,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737536549300527e-05,
+      "loss": 4.023,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736697954549475e-05,
+      "loss": 4.029,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735859359798423e-05,
+      "loss": 4.0318,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735020765047371e-05,
+      "loss": 4.0265,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7341821702963186e-05,
+      "loss": 4.0494,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7333435755452666e-05,
+      "loss": 4.0477,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7325049807942146e-05,
+      "loss": 4.0356,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731668023923536e-05,
+      "loss": 4.0363,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7308294291724835e-05,
+      "loss": 4.0389,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7299908344214315e-05,
+      "loss": 4.0262,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7291522396703795e-05,
+      "loss": 4.0345,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728313644919328e-05,
+      "loss": 4.0337,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727475050168276e-05,
+      "loss": 4.0183,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726636455417224e-05,
+      "loss": 4.0289,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725799498546545e-05,
+      "loss": 4.0264,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724960903795493e-05,
+      "loss": 4.0366,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724122309044441e-05,
+      "loss": 4.0347,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723283714293389e-05,
+      "loss": 4.0343,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722445119542337e-05,
+      "loss": 4.0382,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721606524791285e-05,
+      "loss": 4.0253,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720767930040233e-05,
+      "loss": 4.0398,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719929335289181e-05,
+      "loss": 4.043,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719092378418502e-05,
+      "loss": 4.0242,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71825378366745e-05,
+      "loss": 4.0281,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174168267967716e-05,
+      "loss": 4.0299,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165782320457196e-05,
+      "loss": 4.0236,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157396372946676e-05,
+      "loss": 4.0428,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7149010425436155e-05,
+      "loss": 4.0362,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140624477925635e-05,
+      "loss": 4.0305,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132238530415115e-05,
+      "loss": 4.0348,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123852582904595e-05,
+      "loss": 4.0317,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115466635394075e-05,
+      "loss": 4.0349,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7107097066687284e-05,
+      "loss": 4.0352,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7098711119176764e-05,
+      "loss": 4.0039,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7090325171666244e-05,
+      "loss": 4.0411,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7081939224155724e-05,
+      "loss": 4.0263,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707356965544893e-05,
+      "loss": 4.0407,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706518370793842e-05,
+      "loss": 4.0293,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70567977604279e-05,
+      "loss": 4.0253,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704841181291738e-05,
+      "loss": 4.0214,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704004224421059e-05,
+      "loss": 4.0326,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703165629670007e-05,
+      "loss": 4.0322,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702327034918955e-05,
+      "loss": 4.0294,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701488440167902e-05,
+      "loss": 4.0475,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70064984541685e-05,
+      "loss": 4.0321,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699811250665798e-05,
+      "loss": 4.0064,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698972655914746e-05,
+      "loss": 4.0443,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698134061163694e-05,
+      "loss": 4.019,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697297104293016e-05,
+      "loss": 4.0209,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696458509541964e-05,
+      "loss": 4.0328,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695619914790912e-05,
+      "loss": 4.0328,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69478132003986e-05,
+      "loss": 4.0185,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693944363169181e-05,
+      "loss": 4.0209,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693105768418129e-05,
+      "loss": 4.0148,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922688115474496e-05,
+      "loss": 4.025,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6914302167963976e-05,
+      "loss": 4.036,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905916220453456e-05,
+      "loss": 4.0274,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897530272942936e-05,
+      "loss": 4.0294,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889144325432416e-05,
+      "loss": 4.0219,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880758377921896e-05,
+      "loss": 4.0451,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687238880921511e-05,
+      "loss": 4.0235,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686400286170459e-05,
+      "loss": 4.0347,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685561691419407e-05,
+      "loss": 4.024,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684723096668355e-05,
+      "loss": 4.0273,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683884501917303e-05,
+      "loss": 4.0304,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683045907166251e-05,
+      "loss": 4.0327,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682207312415199e-05,
+      "loss": 4.023,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681368717664147e-05,
+      "loss": 4.0374,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680531760793468e-05,
+      "loss": 4.0178,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679694803922789e-05,
+      "loss": 4.0257,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678856209171737e-05,
+      "loss": 4.0272,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678017614420685e-05,
+      "loss": 4.0319,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677179019669633e-05,
+      "loss": 4.0213,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6763420627989545e-05,
+      "loss": 4.0239,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6755034680479025e-05,
+      "loss": 4.02,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746648732968505e-05,
+      "loss": 4.0272,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738262785457985e-05,
+      "loss": 4.0309,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729876837947465e-05,
+      "loss": 4.0151,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721507269240674e-05,
+      "loss": 4.0299,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713121321730154e-05,
+      "loss": 4.0311,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704735374219634e-05,
+      "loss": 4.0329,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696349426709114e-05,
+      "loss": 4.0063,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687963479198594e-05,
+      "loss": 4.0231,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6679577531688074e-05,
+      "loss": 4.0196,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671191584177554e-05,
+      "loss": 4.0316,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662805636667034e-05,
+      "loss": 4.0338,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665443606796024e-05,
+      "loss": 4.0231,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664606649925346e-05,
+      "loss": 4.0207,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663768055174294e-05,
+      "loss": 4.0229,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662929460423242e-05,
+      "loss": 4.015,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66209086567219e-05,
+      "loss": 4.0224,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661252270921138e-05,
+      "loss": 4.018,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660413676170086e-05,
+      "loss": 4.0262,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659576719299407e-05,
+      "loss": 4.0126,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658738124548355e-05,
+      "loss": 4.0124,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657899529797303e-05,
+      "loss": 4.0145,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657060935046251e-05,
+      "loss": 4.018,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656222340295198e-05,
+      "loss": 4.0227,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655383745544147e-05,
+      "loss": 4.0138,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654545150793095e-05,
+      "loss": 4.0246,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653706556042043e-05,
+      "loss": 4.025,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6528695991713636e-05,
+      "loss": 4.0178,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6520310044203116e-05,
+      "loss": 4.025,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6511924096692596e-05,
+      "loss": 4.0192,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6503538149182076e-05,
+      "loss": 4.0243,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6495168580475285e-05,
+      "loss": 4.0213,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486782632964765e-05,
+      "loss": 4.0157,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478396685454245e-05,
+      "loss": 4.0224,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6470010737943725e-05,
+      "loss": 4.0153,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461641169236934e-05,
+      "loss": 4.0264,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645325522172642e-05,
+      "loss": 4.0243,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64448692742159e-05,
+      "loss": 4.0121,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643648332670538e-05,
+      "loss": 4.018,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642811375799859e-05,
+      "loss": 4.0081,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641972781048807e-05,
+      "loss": 4.0322,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641134186297755e-05,
+      "loss": 4.0136,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640295591546703e-05,
+      "loss": 4.0278,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639456996795651e-05,
+      "loss": 4.0142,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638620039924972e-05,
+      "loss": 4.029,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63778144517392e-05,
+      "loss": 4.028,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636942850422868e-05,
+      "loss": 4.0131,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636104255671816e-05,
+      "loss": 4.0241,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635265660920764e-05,
+      "loss": 4.0146,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634427066169712e-05,
+      "loss": 4.0294,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63358847141866e-05,
+      "loss": 4.0287,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6327498766676085e-05,
+      "loss": 4.0119,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319129197969295e-05,
+      "loss": 4.0186,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310743250458775e-05,
+      "loss": 4.0208,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6302357302948254e-05,
+      "loss": 4.0275,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293971355437734e-05,
+      "loss": 4.0143,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6285585407927214e-05,
+      "loss": 4.026,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627723221802415e-05,
+      "loss": 4.0151,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626884627051363e-05,
+      "loss": 4.0265,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626046032300311e-05,
+      "loss": 4.0202,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1084394454956055,
+      "eval_runtime": 279.5314,
+      "eval_samples_per_second": 1365.11,
+      "eval_steps_per_second": 42.661,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625207437549259e-05,
+      "loss": 4.0171,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624368842798207e-05,
+      "loss": 4.0178,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623531885927529e-05,
+      "loss": 4.0229,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622693291176477e-05,
+      "loss": 4.0252,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621854696425425e-05,
+      "loss": 4.0289,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621016101674373e-05,
+      "loss": 4.0151,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620177506923321e-05,
+      "loss": 4.0155,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619338912172269e-05,
+      "loss": 4.0157,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618500317421217e-05,
+      "loss": 4.0199,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617661722670164e-05,
+      "loss": 4.0252,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616824765799486e-05,
+      "loss": 4.0161,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6159878089288066e-05,
+      "loss": 4.0173,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6151508520581275e-05,
+      "loss": 4.0284,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614312257307076e-05,
+      "loss": 4.0099,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613473662556024e-05,
+      "loss": 4.0106,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612635067804972e-05,
+      "loss": 4.0059,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61179647305392e-05,
+      "loss": 4.0076,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610957878302868e-05,
+      "loss": 4.0137,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610119283551816e-05,
+      "loss": 4.0117,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609280688800764e-05,
+      "loss": 4.0243,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6084420940497115e-05,
+      "loss": 4.028,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607605137179033e-05,
+      "loss": 4.0168,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606766542427981e-05,
+      "loss": 4.0174,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6059279476769284e-05,
+      "loss": 4.0213,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6050893529258764e-05,
+      "loss": 4.003,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6042507581748244e-05,
+      "loss": 4.0191,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603412163423773e-05,
+      "loss": 4.0137,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602573568672721e-05,
+      "loss": 4.0012,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601736611802042e-05,
+      "loss": 4.0102,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60089801705099e-05,
+      "loss": 4.0101,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600059422299938e-05,
+      "loss": 4.014,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599220827548886e-05,
+      "loss": 4.0129,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598382232797834e-05,
+      "loss": 4.0203,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597543638046782e-05,
+      "loss": 4.0154,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59670504329573e-05,
+      "loss": 4.0081,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595866448544678e-05,
+      "loss": 4.0212,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595029491673999e-05,
+      "loss": 4.0245,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594190896922947e-05,
+      "loss": 4.0089,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593352302171895e-05,
+      "loss": 4.0105,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592513707420843e-05,
+      "loss": 4.0117,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5916767505501644e-05,
+      "loss": 4.0028,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908381557991124e-05,
+      "loss": 4.0237,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899995610480604e-05,
+      "loss": 4.0178,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891609662970084e-05,
+      "loss": 4.0201,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5883223715459564e-05,
+      "loss": 4.0122,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587485414675277e-05,
+      "loss": 4.0165,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586646819924225e-05,
+      "loss": 4.015,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585808225173173e-05,
+      "loss": 4.0149,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584969630422121e-05,
+      "loss": 3.989,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584131035671069e-05,
+      "loss": 4.0223,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583292440920017e-05,
+      "loss": 4.0079,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582455484049338e-05,
+      "loss": 4.0248,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581616889298287e-05,
+      "loss": 4.0096,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580778294547235e-05,
+      "loss": 4.0088,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579939699796183e-05,
+      "loss": 4.0024,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57910110504513e-05,
+      "loss": 4.01,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578262510294078e-05,
+      "loss": 4.0193,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577423915543026e-05,
+      "loss": 4.0076,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576585320791974e-05,
+      "loss": 4.0275,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575748363921295e-05,
+      "loss": 4.0113,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574911407050617e-05,
+      "loss": 3.9923,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5740728122995647e-05,
+      "loss": 4.0244,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573234217548512e-05,
+      "loss": 4.003,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5723956227974606e-05,
+      "loss": 4.0028,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571558665926782e-05,
+      "loss": 4.0126,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57072007117573e-05,
+      "loss": 4.0169,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5698814764246776e-05,
+      "loss": 3.9978,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690428816736255e-05,
+      "loss": 4.0056,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5682042869225735e-05,
+      "loss": 3.9955,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5673656921715215e-05,
+      "loss": 4.0118,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5665270974204695e-05,
+      "loss": 4.0141,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5656885026694175e-05,
+      "loss": 4.0083,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5648499079183655e-05,
+      "loss": 4.0129,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5640113131673135e-05,
+      "loss": 4.0015,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5631727184162615e-05,
+      "loss": 4.0242,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5623357615455824e-05,
+      "loss": 4.0111,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5614971667945304e-05,
+      "loss": 4.0178,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5606585720434784e-05,
+      "loss": 4.0064,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559819977292427e-05,
+      "loss": 4.0104,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558981382541375e-05,
+      "loss": 4.0129,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558144425670696e-05,
+      "loss": 4.0138,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557305830919644e-05,
+      "loss": 4.0082,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556467236168592e-05,
+      "loss": 4.0182,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55562864141754e-05,
+      "loss": 3.9983,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554790046666488e-05,
+      "loss": 4.0072,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553951451915436e-05,
+      "loss": 4.0074,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553114495044757e-05,
+      "loss": 4.018,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552275900293705e-05,
+      "loss": 4.0014,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551437305542653e-05,
+      "loss": 4.0048,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550598710791601e-05,
+      "loss": 4.0067,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549760116040549e-05,
+      "loss": 4.0089,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548921521289497e-05,
+      "loss": 4.011,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548082926538445e-05,
+      "loss": 3.9988,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5472459696677664e-05,
+      "loss": 4.0113,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546407374916714e-05,
+      "loss": 4.0135,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545568780165662e-05,
+      "loss": 4.013,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54473018541461e-05,
+      "loss": 3.9903,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543891590663558e-05,
+      "loss": 4.0032,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543052995912506e-05,
+      "loss": 4.0013,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542214401161454e-05,
+      "loss": 4.0146,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541375806410402e-05,
+      "loss": 4.0158,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5405388495397226e-05,
+      "loss": 4.0052,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5397002547886706e-05,
+      "loss": 4.0024,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538861660037619e-05,
+      "loss": 4.0052,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538023065286567e-05,
+      "loss": 3.9966,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537184470535515e-05,
+      "loss": 4.0063,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536347513664836e-05,
+      "loss": 4.0005,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535508918913784e-05,
+      "loss": 4.0083,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534670324162732e-05,
+      "loss": 3.9986,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53383172941168e-05,
+      "loss": 3.9927,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532994772541001e-05,
+      "loss": 3.9979,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532156177789949e-05,
+      "loss": 3.9966,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531317583038897e-05,
+      "loss": 4.0059,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530478988287845e-05,
+      "loss": 3.9969,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529640393536793e-05,
+      "loss": 4.0074,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528803436666115e-05,
+      "loss": 4.0054,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527964841915063e-05,
+      "loss": 4.004,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5271278850443836e-05,
+      "loss": 4.0062,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5262892902933316e-05,
+      "loss": 3.9992,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254506955422796e-05,
+      "loss": 4.0091,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5246137386716005e-05,
+      "loss": 4.0028,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237751439205485e-05,
+      "loss": 3.9992,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5229365491694965e-05,
+      "loss": 4.0042,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5220979544184445e-05,
+      "loss": 3.9993,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5212593596673925e-05,
+      "loss": 4.0075,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5204207649163405e-05,
+      "loss": 4.0018,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5195821701652885e-05,
+      "loss": 3.9959,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51874521329461e-05,
+      "loss": 3.9983,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517906618543558e-05,
+      "loss": 3.9931,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517068023792506e-05,
+      "loss": 4.0119,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516229429041454e-05,
+      "loss": 4.001,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515390834290402e-05,
+      "loss": 4.0113,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145522395393493e-05,
+      "loss": 3.9954,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137136447882973e-05,
+      "loss": 4.009,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512875050037245e-05,
+      "loss": 4.0092,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512036455286193e-05,
+      "loss": 4.0005,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511197860535141e-05,
+      "loss": 4.0032,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510359265784089e-05,
+      "loss": 4.0005,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509520671033037e-05,
+      "loss": 4.0119,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508683714162358e-05,
+      "loss": 4.0127,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50784675729168e-05,
+      "loss": 3.9974,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507008162540628e-05,
+      "loss": 3.9965,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506169567789576e-05,
+      "loss": 4.0064,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505330973038524e-05,
+      "loss": 4.0087,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504494016167845e-05,
+      "loss": 3.995,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503655421416793e-05,
+      "loss": 4.008,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502816826665741e-05,
+      "loss": 3.9977,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501978231914689e-05,
+      "loss": 4.0098,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501139637163637e-05,
+      "loss": 4.0025,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0972418785095215,
+      "eval_runtime": 280.2207,
+      "eval_samples_per_second": 1361.752,
+      "eval_steps_per_second": 42.556,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500301042412585e-05,
+      "loss": 4.0111,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499462447661533e-05,
+      "loss": 3.9949,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498623852910481e-05,
+      "loss": 4.0046,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497785258159429e-05,
+      "loss": 4.0104,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496946663408377e-05,
+      "loss": 4.0097,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496108068657325e-05,
+      "loss": 4.0019,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952694739062733e-05,
+      "loss": 3.9981,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494430879155221e-05,
+      "loss": 3.9987,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4935922844041687e-05,
+      "loss": 4.0,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4927536896531166e-05,
+      "loss": 4.0044,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4919150949020646e-05,
+      "loss": 4.0013,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4910765001510126e-05,
+      "loss": 4.0014,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4902395432803336e-05,
+      "loss": 4.0137,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4894009485292815e-05,
+      "loss": 3.9908,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4885623537782295e-05,
+      "loss": 3.9942,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4877237590271775e-05,
+      "loss": 3.9895,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4868851642761255e-05,
+      "loss": 3.9892,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860465695250735e-05,
+      "loss": 3.9971,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4852079747740215e-05,
+      "loss": 3.9948,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48436938002297e-05,
+      "loss": 4.0047,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483530785271918e-05,
+      "loss": 4.0134,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482692190520866e-05,
+      "loss": 4.0006,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481853595769814e-05,
+      "loss": 4.003,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481015001018762e-05,
+      "loss": 4.0038,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480178044148083e-05,
+      "loss": 3.9853,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479339449397031e-05,
+      "loss": 4.0067,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478500854645979e-05,
+      "loss": 3.9948,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477662259894927e-05,
+      "loss": 3.9877,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476825303024248e-05,
+      "loss": 3.9951,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475986708273196e-05,
+      "loss": 3.9907,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475149751402517e-05,
+      "loss": 3.9955,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4743111566514656e-05,
+      "loss": 3.9932,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734725619004136e-05,
+      "loss": 4.0067,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726339671493616e-05,
+      "loss": 3.9972,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717953723983095e-05,
+      "loss": 3.9931,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709584155276305e-05,
+      "loss": 4.0022,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4701198207765785e-05,
+      "loss": 4.008,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4692812260255265e-05,
+      "loss": 3.9929,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4684426312744744e-05,
+      "loss": 3.9912,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467604036523422e-05,
+      "loss": 3.9944,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4667670796527434e-05,
+      "loss": 3.9912,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4659284849016913e-05,
+      "loss": 4.0017,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4650898901506393e-05,
+      "loss": 4.0004,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464251295399587e-05,
+      "loss": 4.0064,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463412700648535e-05,
+      "loss": 3.9976,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462574105897483e-05,
+      "loss": 3.9977,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461735511146431e-05,
+      "loss": 3.9993,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460898554275752e-05,
+      "loss": 3.9973,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460061597405074e-05,
+      "loss": 3.9723,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459223002654022e-05,
+      "loss": 4.003,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458384407902969e-05,
+      "loss": 3.9947,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457545813151917e-05,
+      "loss": 4.0079,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456707218400865e-05,
+      "loss": 3.991,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455868623649813e-05,
+      "loss": 3.9958,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455030028898761e-05,
+      "loss": 3.983,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541947099084556e-05,
+      "loss": 3.9901,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453356115157404e-05,
+      "loss": 4.0092,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525175204063516e-05,
+      "loss": 3.9923,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516789256552996e-05,
+      "loss": 4.0096,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508403309042476e-05,
+      "loss": 3.9962,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4500017361531956e-05,
+      "loss": 3.9731,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4491631414021436e-05,
+      "loss": 4.0122,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4483245466510916e-05,
+      "loss": 3.9839,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4474859519000396e-05,
+      "loss": 3.9852,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4466489950293605e-05,
+      "loss": 3.9993,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4458104002783085e-05,
+      "loss": 3.9985,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4449718055272565e-05,
+      "loss": 3.9851,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4441332107762045e-05,
+      "loss": 3.988,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4432946160251525e-05,
+      "loss": 3.9815,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442456021274101e-05,
+      "loss": 3.9877,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441617426523049e-05,
+      "loss": 4.0018,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440778831771997e-05,
+      "loss": 3.9937,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439940237020945e-05,
+      "loss": 3.9934,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439101642269893e-05,
+      "loss": 3.9859,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438266323279587e-05,
+      "loss": 4.0064,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437427728528535e-05,
+      "loss": 3.9936,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436589133777483e-05,
+      "loss": 3.9991,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435750539026431e-05,
+      "loss": 3.9892,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434911944275379e-05,
+      "loss": 3.9934,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340749874047e-05,
+      "loss": 3.999,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433236392653648e-05,
+      "loss": 3.9947,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4323977979025965e-05,
+      "loss": 3.9921,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315592031515445e-05,
+      "loss": 4.0019,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307206084004925e-05,
+      "loss": 3.9844,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298820136494405e-05,
+      "loss": 3.9882,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429043418898388e-05,
+      "loss": 3.9931,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428204824147336e-05,
+      "loss": 4.0004,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427366229396284e-05,
+      "loss": 3.9865,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265292725256054e-05,
+      "loss": 3.9929,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425690677774553e-05,
+      "loss": 3.9869,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424852083023501e-05,
+      "loss": 3.9911,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424013488272449e-05,
+      "loss": 3.9929,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423174893521397e-05,
+      "loss": 3.986,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422337936650718e-05,
+      "loss": 3.9942,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421499341899666e-05,
+      "loss": 3.9988,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420660747148614e-05,
+      "loss": 4.0024,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419822152397562e-05,
+      "loss": 3.9695,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418985195526883e-05,
+      "loss": 3.9897,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418146600775831e-05,
+      "loss": 3.9827,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417308006024779e-05,
+      "loss": 3.9992,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416469411273727e-05,
+      "loss": 4.0018,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415630816522675e-05,
+      "loss": 3.988,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414793859651996e-05,
+      "loss": 3.9859,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413956902781318e-05,
+      "loss": 3.9896,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131183080302657e-05,
+      "loss": 3.9836,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122797132792137e-05,
+      "loss": 3.9861,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114411185281617e-05,
+      "loss": 3.9829,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4106025237771096e-05,
+      "loss": 3.991,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097639290260576e-05,
+      "loss": 3.9816,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4089253342750056e-05,
+      "loss": 3.9796,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4080867395239536e-05,
+      "loss": 3.9832,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4072497826532745e-05,
+      "loss": 3.9782,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4064128257825955e-05,
+      "loss": 3.9905,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4055742310315434e-05,
+      "loss": 3.9776,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047356362804914e-05,
+      "loss": 3.9929,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4038970415294394e-05,
+      "loss": 3.9892,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4030584467783874e-05,
+      "loss": 3.9891,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4022198520273354e-05,
+      "loss": 3.9883,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4013812572762834e-05,
+      "loss": 3.9857,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400542662525232e-05,
+      "loss": 3.991,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399705705654553e-05,
+      "loss": 3.9848,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398867110903501e-05,
+      "loss": 3.9812,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398030154032822e-05,
+      "loss": 3.9897,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39719155928177e-05,
+      "loss": 3.9857,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396352964530718e-05,
+      "loss": 3.9898,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395514369779666e-05,
+      "loss": 3.9848,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394675775028614e-05,
+      "loss": 3.983,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393838818157935e-05,
+      "loss": 3.9853,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393000223406883e-05,
+      "loss": 3.979,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392161628655831e-05,
+      "loss": 3.9908,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391323033904779e-05,
+      "loss": 3.9851,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904844391537275e-05,
+      "loss": 3.9978,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896458444026755e-05,
+      "loss": 3.9798,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888072496516235e-05,
+      "loss": 3.9893,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879686549005715e-05,
+      "loss": 3.9965,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387130060149519e-05,
+      "loss": 3.987,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862931032788404e-05,
+      "loss": 3.9821,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854545085277884e-05,
+      "loss": 3.986,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3846159137767363e-05,
+      "loss": 3.9964,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383777319025684e-05,
+      "loss": 3.9951,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382938724274632e-05,
+      "loss": 3.9846,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382101767403953e-05,
+      "loss": 3.9816,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381263172652901e-05,
+      "loss": 3.9892,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380424577901849e-05,
+      "loss": 3.9942,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379585983150797e-05,
+      "loss": 3.9747,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378747388399745e-05,
+      "loss": 3.9994,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377910431529066e-05,
+      "loss": 3.9796,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377071836778014e-05,
+      "loss": 3.994,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376233242026962e-05,
+      "loss": 3.989,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.088191986083984,
+      "eval_runtime": 292.4769,
+      "eval_samples_per_second": 1304.688,
+      "eval_steps_per_second": 40.772,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.37539464727591e-05,
+      "loss": 3.9883,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374557690405231e-05,
+      "loss": 3.9805,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373719095654179e-05,
+      "loss": 3.9859,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372880500903127e-05,
+      "loss": 3.9973,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372041906152075e-05,
+      "loss": 3.9874,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371203311401023e-05,
+      "loss": 3.9905,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703663545303446e-05,
+      "loss": 3.9811,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369529397659666e-05,
+      "loss": 3.9805,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3686908029086135e-05,
+      "loss": 3.9853,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367853846037935e-05,
+      "loss": 3.9921,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367015251286883e-05,
+      "loss": 3.9825,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366176656535831e-05,
+      "loss": 3.9913,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365339699665152e-05,
+      "loss": 3.9948,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3645011049141e-05,
+      "loss": 3.9745,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363662510163048e-05,
+      "loss": 3.982,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362823915411995e-05,
+      "loss": 3.9707,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361985320660943e-05,
+      "loss": 3.9709,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361146725909892e-05,
+      "loss": 3.9842,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36030813115884e-05,
+      "loss": 3.9755,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359469536407788e-05,
+      "loss": 3.9885,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358630941656736e-05,
+      "loss": 3.9978,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357792346905684e-05,
+      "loss": 3.9834,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356953752154632e-05,
+      "loss": 3.9904,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35611515740358e-05,
+      "loss": 3.9887,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355276562652528e-05,
+      "loss": 3.9704,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354439605781849e-05,
+      "loss": 3.988,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353601011030797e-05,
+      "loss": 3.9797,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352762416279745e-05,
+      "loss": 3.9749,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351923821528693e-05,
+      "loss": 3.9796,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351085226777641e-05,
+      "loss": 3.9719,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350246632026589e-05,
+      "loss": 3.981,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349408037275537e-05,
+      "loss": 3.9824,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348569442524485e-05,
+      "loss": 3.9884,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477324856538064e-05,
+      "loss": 3.9812,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468938909027544e-05,
+      "loss": 3.9786,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346056934032075e-05,
+      "loss": 3.9798,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345218339281023e-05,
+      "loss": 3.9949,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344381382410344e-05,
+      "loss": 3.9781,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343542787659292e-05,
+      "loss": 3.9775,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34270419290824e-05,
+      "loss": 3.9771,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341865598157188e-05,
+      "loss": 3.9799,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341027003406136e-05,
+      "loss": 3.9808,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340188408655084e-05,
+      "loss": 3.9884,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339349813904032e-05,
+      "loss": 3.9882,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33851121915298e-05,
+      "loss": 3.98,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337672624401928e-05,
+      "loss": 3.9815,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336834029650876e-05,
+      "loss": 3.9807,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335995434899824e-05,
+      "loss": 3.9824,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335158478029145e-05,
+      "loss": 3.9638,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334319883278093e-05,
+      "loss": 3.9852,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333482926407414e-05,
+      "loss": 3.9802,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332644331656362e-05,
+      "loss": 3.9907,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33180573690531e-05,
+      "loss": 3.9761,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330967142154258e-05,
+      "loss": 3.9782,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330128547403206e-05,
+      "loss": 3.9735,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292915905325276e-05,
+      "loss": 3.9704,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284529957814756e-05,
+      "loss": 3.9927,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3276144010304236e-05,
+      "loss": 3.9749,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3267758062793716e-05,
+      "loss": 4.0013,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3259372115283195e-05,
+      "loss": 3.9771,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3251002546576405e-05,
+      "loss": 3.9639,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3242616599065885e-05,
+      "loss": 3.9956,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3234230651555364e-05,
+      "loss": 3.9734,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3225844704044844e-05,
+      "loss": 3.9675,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3217458756534324e-05,
+      "loss": 3.977,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3209072809023804e-05,
+      "loss": 3.9871,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3200703240317013e-05,
+      "loss": 3.9663,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3192317292806493e-05,
+      "loss": 3.9732,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318393134529597e-05,
+      "loss": 3.9665,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317554539778546e-05,
+      "loss": 3.9696,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316715945027494e-05,
+      "loss": 3.9907,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315877350276442e-05,
+      "loss": 3.9784,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31503875552539e-05,
+      "loss": 3.9796,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314200160774338e-05,
+      "loss": 3.972,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313363203903659e-05,
+      "loss": 3.99,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312524609152607e-05,
+      "loss": 3.9803,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311686014401555e-05,
+      "loss": 3.9863,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310850695411249e-05,
+      "loss": 3.9738,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310012100660197e-05,
+      "loss": 3.9769,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309173505909145e-05,
+      "loss": 3.9841,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308334911158093e-05,
+      "loss": 3.9783,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3074963164070414e-05,
+      "loss": 3.9778,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3066577216559894e-05,
+      "loss": 3.9855,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058191269049374e-05,
+      "loss": 3.9774,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049805321538854e-05,
+      "loss": 3.9696,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3041419374028334e-05,
+      "loss": 3.9773,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303303342651781e-05,
+      "loss": 3.9833,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302464747900729e-05,
+      "loss": 3.9746,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301626153149677e-05,
+      "loss": 3.9757,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3007891962789976e-05,
+      "loss": 3.9701,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2999506015279456e-05,
+      "loss": 3.975,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2991120067768936e-05,
+      "loss": 3.9769,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2982734120258416e-05,
+      "loss": 3.9742,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2974348172747896e-05,
+      "loss": 3.9789,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296597860404111e-05,
+      "loss": 3.981,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295759265653059e-05,
+      "loss": 3.9876,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294922308782381e-05,
+      "loss": 3.9562,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294083714031328e-05,
+      "loss": 3.9738,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293245119280276e-05,
+      "loss": 3.9708,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292406524529224e-05,
+      "loss": 3.9803,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291567929778172e-05,
+      "loss": 3.986,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29072933502712e-05,
+      "loss": 3.9715,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289890740276068e-05,
+      "loss": 3.9706,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289052145525016e-05,
+      "loss": 3.9743,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288215188654337e-05,
+      "loss": 3.9716,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287376593903285e-05,
+      "loss": 3.9693,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2865396370326065e-05,
+      "loss": 3.9693,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2857010422815545e-05,
+      "loss": 3.9759,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2848624475305025e-05,
+      "loss": 3.9663,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2840238527794505e-05,
+      "loss": 3.9661,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2831852580283985e-05,
+      "loss": 3.9709,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823466632773465e-05,
+      "loss": 3.9601,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2815097064066674e-05,
+      "loss": 3.976,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2806711116556154e-05,
+      "loss": 3.9692,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2798325169045634e-05,
+      "loss": 3.9716,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2789939221535114e-05,
+      "loss": 3.9757,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2781553274024594e-05,
+      "loss": 3.9768,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2773167326514074e-05,
+      "loss": 3.9729,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2764781379003554e-05,
+      "loss": 3.9693,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2756395431493034e-05,
+      "loss": 3.9797,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2748009483982514e-05,
+      "loss": 3.9701,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273965629407946e-05,
+      "loss": 3.9653,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273127034656894e-05,
+      "loss": 3.9761,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272288439905842e-05,
+      "loss": 3.9722,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27144984515479e-05,
+      "loss": 3.9731,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270611250403738e-05,
+      "loss": 3.9696,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269772655652686e-05,
+      "loss": 3.9735,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268935698782007e-05,
+      "loss": 3.9703,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268097104030955e-05,
+      "loss": 3.9664,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267258509279903e-05,
+      "loss": 3.9674,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266419914528851e-05,
+      "loss": 3.9791,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265581319777799e-05,
+      "loss": 3.9849,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264742725026747e-05,
+      "loss": 3.9629,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263904130275695e-05,
+      "loss": 3.9711,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263065535524643e-05,
+      "loss": 3.984,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262226940773591e-05,
+      "loss": 3.973,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2613899839029116e-05,
+      "loss": 3.9688,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260553027032233e-05,
+      "loss": 3.9707,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259714432281181e-05,
+      "loss": 3.9835,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258877475410502e-05,
+      "loss": 3.9816,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25803888065945e-05,
+      "loss": 3.969,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257200285908398e-05,
+      "loss": 3.9645,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256361691157346e-05,
+      "loss": 3.9768,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255523096406294e-05,
+      "loss": 3.9771,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254684501655242e-05,
+      "loss": 3.9634,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25384590690419e-05,
+      "loss": 3.9835,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253007312153138e-05,
+      "loss": 3.9691,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252168717402086e-05,
+      "loss": 3.9787,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251330122651034e-05,
+      "loss": 3.9751,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.079358100891113,
+      "eval_runtime": 290.5247,
+      "eval_samples_per_second": 1313.454,
+      "eval_steps_per_second": 41.046,
+      "step": 1068480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.250491527899982e-05,
+      "loss": 3.9744,
+      "step": 1068544
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.24965293314893e-05,
+      "loss": 3.964,
+      "step": 1069056
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.248815976278251e-05,
+      "loss": 3.9736,
+      "step": 1069568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.247977381527199e-05,
+      "loss": 3.9843,
+      "step": 1070080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.247138786776147e-05,
+      "loss": 3.9733,
+      "step": 1070592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.246300192025095e-05,
+      "loss": 3.9739,
+      "step": 1071104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.245463235154416e-05,
+      "loss": 3.9712,
+      "step": 1071616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2446246404033646e-05,
+      "loss": 3.9674,
+      "step": 1072128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2437860456523125e-05,
+      "loss": 3.9689,
+      "step": 1072640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2429474509012605e-05,
+      "loss": 3.9775,
+      "step": 1073152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2421104940305815e-05,
+      "loss": 3.9698,
+      "step": 1073664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2412718992795294e-05,
+      "loss": 3.9784,
+      "step": 1074176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2404349424088504e-05,
+      "loss": 3.9815,
+      "step": 1074688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2395963476577984e-05,
+      "loss": 3.9596,
+      "step": 1075200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2387577529067463e-05,
+      "loss": 3.9692,
+      "step": 1075712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2379191581556943e-05,
+      "loss": 3.9556,
+      "step": 1076224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.237082201285015e-05,
+      "loss": 3.9608,
+      "step": 1076736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.236243606533963e-05,
+      "loss": 3.9688,
+      "step": 1077248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.235405011782911e-05,
+      "loss": 3.961,
+      "step": 1077760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.23456641703186e-05,
+      "loss": 3.9731,
+      "step": 1078272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.233727822280808e-05,
+      "loss": 3.9855,
+      "step": 1078784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.232889227529756e-05,
+      "loss": 3.966,
+      "step": 1079296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.232050632778704e-05,
+      "loss": 3.9754,
+      "step": 1079808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.231212038027652e-05,
+      "loss": 3.9766,
+      "step": 1080320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2303734432766e-05,
+      "loss": 3.9583,
+      "step": 1080832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.229536486405921e-05,
+      "loss": 3.9778,
+      "step": 1081344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.228697891654869e-05,
+      "loss": 3.9621,
+      "step": 1081856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.227859296903817e-05,
+      "loss": 3.9643,
+      "step": 1082368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.227020702152765e-05,
+      "loss": 3.9649,
+      "step": 1082880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.226182107401712e-05,
+      "loss": 3.9571,
+      "step": 1083392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.22534351265066e-05,
+      "loss": 3.9656,
+      "step": 1083904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.224504917899608e-05,
+      "loss": 3.9707,
+      "step": 1084416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.223666323148556e-05,
+      "loss": 3.9767,
+      "step": 1084928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.222829366277878e-05,
+      "loss": 3.9693,
+      "step": 1085440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.221990771526826e-05,
+      "loss": 3.9611,
+      "step": 1085952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.221152176775774e-05,
+      "loss": 3.9661,
+      "step": 1086464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.220313582024722e-05,
+      "loss": 3.9842,
+      "step": 1086976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2194766251540426e-05,
+      "loss": 3.9619,
+      "step": 1087488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2186380304029906e-05,
+      "loss": 3.9664,
+      "step": 1088000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.217801073532312e-05,
+      "loss": 3.9656,
+      "step": 1088512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2169624787812595e-05,
+      "loss": 3.9617,
+      "step": 1089024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2161238840302075e-05,
+      "loss": 3.9692,
+      "step": 1089536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2152852892791555e-05,
+      "loss": 3.9741,
+      "step": 1090048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2144466945281035e-05,
+      "loss": 3.9785,
+      "step": 1090560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2136080997770515e-05,
+      "loss": 3.968,
+      "step": 1091072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.212769505026e-05,
+      "loss": 3.9659,
+      "step": 1091584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.211930910274948e-05,
+      "loss": 3.9661,
+      "step": 1092096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.211092315523896e-05,
+      "loss": 3.9735,
+      "step": 1092608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.210255358653217e-05,
+      "loss": 3.9475,
+      "step": 1093120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.209416763902165e-05,
+      "loss": 3.9682,
+      "step": 1093632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.208578169151113e-05,
+      "loss": 3.9721,
+      "step": 1094144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.207739574400061e-05,
+      "loss": 3.9764,
+      "step": 1094656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.206902617529382e-05,
+      "loss": 3.9635,
+      "step": 1095168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.20606402277833e-05,
+      "loss": 3.9592,
+      "step": 1095680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.205225428027278e-05,
+      "loss": 3.9653,
+      "step": 1096192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.204388471156599e-05,
+      "loss": 3.955,
+      "step": 1096704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.203549876405547e-05,
+      "loss": 3.9777,
+      "step": 1097216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2027112816544955e-05,
+      "loss": 3.9634,
+      "step": 1097728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2018726869034435e-05,
+      "loss": 3.9834,
+      "step": 1098240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2010340921523915e-05,
+      "loss": 3.9663,
+      "step": 1098752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2001954974013395e-05,
+      "loss": 3.9495,
+      "step": 1099264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1993569026502875e-05,
+      "loss": 3.9787,
+      "step": 1099776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1985183078992355e-05,
+      "loss": 3.9555,
+      "step": 1100288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1976797131481835e-05,
+      "loss": 3.9564,
+      "step": 1100800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.196841118397131e-05,
+      "loss": 3.9654,
+      "step": 1101312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1960041615264524e-05,
+      "loss": 3.9708,
+      "step": 1101824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.195167204655773e-05,
+      "loss": 3.9535,
+      "step": 1102336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.194328609904721e-05,
+      "loss": 3.9574,
+      "step": 1102848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.193490015153669e-05,
+      "loss": 3.9571,
+      "step": 1103360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.192651420402617e-05,
+      "loss": 3.9565,
+      "step": 1103872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.191812825651565e-05,
+      "loss": 3.976,
+      "step": 1104384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.190974230900513e-05,
+      "loss": 3.9675,
+      "step": 1104896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.190135636149461e-05,
+      "loss": 3.9625,
+      "step": 1105408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.189297041398409e-05,
+      "loss": 3.9607,
+      "step": 1105920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.188461722408104e-05,
+      "loss": 3.9753,
+      "step": 1106432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.187623127657052e-05,
+      "loss": 3.9646,
+      "step": 1106944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.186784532906e-05,
+      "loss": 3.9693,
+      "step": 1107456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.185945938154948e-05,
+      "loss": 3.964,
+      "step": 1107968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.185107343403896e-05,
+      "loss": 3.9659,
+      "step": 1108480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.184268748652843e-05,
+      "loss": 3.969,
+      "step": 1108992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.183430153901791e-05,
+      "loss": 3.9649,
+      "step": 1109504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1825931970311126e-05,
+      "loss": 3.9649,
+      "step": 1110016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1817546022800606e-05,
+      "loss": 3.9716,
+      "step": 1110528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1809160075290086e-05,
+      "loss": 3.9698,
+      "step": 1111040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1800774127779566e-05,
+      "loss": 3.9519,
+      "step": 1111552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1792388180269046e-05,
+      "loss": 3.9634,
+      "step": 1112064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1784002232758526e-05,
+      "loss": 3.9726,
+      "step": 1112576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1775632664051735e-05,
+      "loss": 3.959,
+      "step": 1113088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1767246716541215e-05,
+      "loss": 3.9628,
+      "step": 1113600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1758860769030695e-05,
+      "loss": 3.9586,
+      "step": 1114112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1750474821520175e-05,
+      "loss": 3.9616,
+      "step": 1114624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1742088874009655e-05,
+      "loss": 3.9613,
+      "step": 1115136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1733702926499135e-05,
+      "loss": 3.9616,
+      "step": 1115648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1725316978988615e-05,
+      "loss": 3.9664,
+      "step": 1116160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1716931031478095e-05,
+      "loss": 3.9634,
+      "step": 1116672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.170856146277131e-05,
+      "loss": 3.9769,
+      "step": 1117184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.170017551526079e-05,
+      "loss": 3.9407,
+      "step": 1117696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.169178956775027e-05,
+      "loss": 3.9643,
+      "step": 1118208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.168340362023975e-05,
+      "loss": 3.9579,
+      "step": 1118720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.167503405153296e-05,
+      "loss": 3.9673,
+      "step": 1119232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.166664810402244e-05,
+      "loss": 3.9714,
+      "step": 1119744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.165826215651192e-05,
+      "loss": 3.9612,
+      "step": 1120256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.16498762090014e-05,
+      "loss": 3.9574,
+      "step": 1120768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.164150664029461e-05,
+      "loss": 3.962,
+      "step": 1121280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.163312069278409e-05,
+      "loss": 3.9618,
+      "step": 1121792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.162473474527357e-05,
+      "loss": 3.9547,
+      "step": 1122304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.161634879776305e-05,
+      "loss": 3.9554,
+      "step": 1122816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1607979229056265e-05,
+      "loss": 3.961,
+      "step": 1123328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1599609660349474e-05,
+      "loss": 3.9607,
+      "step": 1123840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1591223712838954e-05,
+      "loss": 3.9504,
+      "step": 1124352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1582837765328434e-05,
+      "loss": 3.9598,
+      "step": 1124864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1574451817817914e-05,
+      "loss": 3.9465,
+      "step": 1125376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1566065870307393e-05,
+      "loss": 3.9668,
+      "step": 1125888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1557679922796873e-05,
+      "loss": 3.9516,
+      "step": 1126400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.154929397528635e-05,
+      "loss": 3.9584,
+      "step": 1126912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.154090802777583e-05,
+      "loss": 3.9623,
+      "step": 1127424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.153252208026531e-05,
+      "loss": 3.9655,
+      "step": 1127936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.152415251155852e-05,
+      "loss": 3.9588,
+      "step": 1128448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1515766564048e-05,
+      "loss": 3.9578,
+      "step": 1128960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.150738061653748e-05,
+      "loss": 3.9639,
+      "step": 1129472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.149899466902696e-05,
+      "loss": 3.9558,
+      "step": 1129984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.149060872151644e-05,
+      "loss": 3.9528,
+      "step": 1130496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.148222277400592e-05,
+      "loss": 3.9624,
+      "step": 1131008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.14738368264954e-05,
+      "loss": 3.9635,
+      "step": 1131520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.146545087898488e-05,
+      "loss": 3.9634,
+      "step": 1132032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.145708131027809e-05,
+      "loss": 3.9562,
+      "step": 1132544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.144869536276757e-05,
+      "loss": 3.958,
+      "step": 1133056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.144030941525705e-05,
+      "loss": 3.9567,
+      "step": 1133568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.143192346774653e-05,
+      "loss": 3.9538,
+      "step": 1134080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.142355389903974e-05,
+      "loss": 3.9508,
+      "step": 1134592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.141516795152922e-05,
+      "loss": 3.9693,
+      "step": 1135104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.14067820040187e-05,
+      "loss": 3.9727,
+      "step": 1135616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.139839605650819e-05,
+      "loss": 3.955,
+      "step": 1136128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1390026487801396e-05,
+      "loss": 3.9576,
+      "step": 1136640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1381640540290876e-05,
+      "loss": 3.9662,
+      "step": 1137152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1373254592780356e-05,
+      "loss": 3.9614,
+      "step": 1137664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1364868645269836e-05,
+      "loss": 3.9566,
+      "step": 1138176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1356499076563045e-05,
+      "loss": 3.9614,
+      "step": 1138688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1348113129052525e-05,
+      "loss": 3.9715,
+      "step": 1139200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1339727181542005e-05,
+      "loss": 3.967,
+      "step": 1139712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1331341234031485e-05,
+      "loss": 3.9576,
+      "step": 1140224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1322971665324694e-05,
+      "loss": 3.9529,
+      "step": 1140736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1314585717814174e-05,
+      "loss": 3.9638,
+      "step": 1141248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1306199770303654e-05,
+      "loss": 3.9642,
+      "step": 1141760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.129781382279314e-05,
+      "loss": 3.9497,
+      "step": 1142272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.128944425408635e-05,
+      "loss": 3.972,
+      "step": 1142784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.128105830657583e-05,
+      "loss": 3.9535,
+      "step": 1143296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.127268873786904e-05,
+      "loss": 3.9647,
+      "step": 1143808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.126430279035852e-05,
+      "loss": 3.9595,
+      "step": 1144320
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.072823524475098,
+      "eval_runtime": 291.5404,
+      "eval_samples_per_second": 1308.879,
+      "eval_steps_per_second": 40.903,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1255916842848e-05,
+      "loss": 3.9588,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124753089533748e-05,
+      "loss": 3.9565,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123914494782696e-05,
+      "loss": 3.9612,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123075900031644e-05,
+      "loss": 3.9686,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122237305280592e-05,
+      "loss": 3.9638,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12139871052954e-05,
+      "loss": 3.9612,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120560115778488e-05,
+      "loss": 3.9572,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119721521027436e-05,
+      "loss": 3.9559,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118882926276384e-05,
+      "loss": 3.9575,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1180443315253325e-05,
+      "loss": 3.9677,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1172057367742805e-05,
+      "loss": 3.9576,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116367142023228e-05,
+      "loss": 3.9617,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1155301851525494e-05,
+      "loss": 3.9704,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1146915904014974e-05,
+      "loss": 3.9494,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1138529956504454e-05,
+      "loss": 3.9557,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113014400899393e-05,
+      "loss": 3.9408,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112175806148341e-05,
+      "loss": 3.9488,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111337211397289e-05,
+      "loss": 3.9564,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110498616646237e-05,
+      "loss": 3.9491,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109660021895185e-05,
+      "loss": 3.9575,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108821427144133e-05,
+      "loss": 3.974,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107982832393081e-05,
+      "loss": 3.9544,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107144237642029e-05,
+      "loss": 3.9616,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1063056428909773e-05,
+      "loss": 3.9632,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105468686020298e-05,
+      "loss": 3.9525,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104630091269246e-05,
+      "loss": 3.9576,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103791496518194e-05,
+      "loss": 3.9459,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102952901767142e-05,
+      "loss": 3.9567,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10211430701609e-05,
+      "loss": 3.9544,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101275712265038e-05,
+      "loss": 3.9461,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100438755394359e-05,
+      "loss": 3.9585,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099600160643307e-05,
+      "loss": 3.9498,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098761565892255e-05,
+      "loss": 3.9627,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097922971141203e-05,
+      "loss": 3.9624,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097086014270524e-05,
+      "loss": 3.9482,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0962490573998456e-05,
+      "loss": 3.9533,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0954104626487936e-05,
+      "loss": 3.9721,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0945735057781145e-05,
+      "loss": 3.9485,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0937349110270625e-05,
+      "loss": 3.9545,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0928963162760105e-05,
+      "loss": 3.952,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0920577215249585e-05,
+      "loss": 3.9507,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0912191267739065e-05,
+      "loss": 3.9517,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0903805320228545e-05,
+      "loss": 3.9686,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0895419372718025e-05,
+      "loss": 3.9621,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0887033425207505e-05,
+      "loss": 3.9553,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0878663856500714e-05,
+      "loss": 3.9579,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0870277908990194e-05,
+      "loss": 3.9485,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086189196147968e-05,
+      "loss": 3.9639,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085350601396916e-05,
+      "loss": 3.935,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084512006645864e-05,
+      "loss": 3.9556,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0836734118948114e-05,
+      "loss": 3.9592,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0828348171437594e-05,
+      "loss": 3.9646,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0819962223927074e-05,
+      "loss": 3.9529,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081159265522028e-05,
+      "loss": 3.9447,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.080320670770976e-05,
+      "loss": 3.9545,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.079482076019924e-05,
+      "loss": 3.9426,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078645119149246e-05,
+      "loss": 3.9708,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077806524398193e-05,
+      "loss": 3.9496,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076967929647142e-05,
+      "loss": 3.9706,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07612933489609e-05,
+      "loss": 3.9518,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075290740145038e-05,
+      "loss": 3.9427,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074452145393986e-05,
+      "loss": 3.9621,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073613550642934e-05,
+      "loss": 3.9433,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072774955891882e-05,
+      "loss": 3.9502,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07193636114083e-05,
+      "loss": 3.9471,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071099404270151e-05,
+      "loss": 3.9606,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070264085279845e-05,
+      "loss": 3.9437,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069425490528793e-05,
+      "loss": 3.9472,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0685868957777405e-05,
+      "loss": 3.9422,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0677483010266885e-05,
+      "loss": 3.9429,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066909706275637e-05,
+      "loss": 3.9635,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066071111524585e-05,
+      "loss": 3.9583,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065232516773533e-05,
+      "loss": 3.9475,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064393922022481e-05,
+      "loss": 3.953,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.063555327271429e-05,
+      "loss": 3.9627,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062716732520377e-05,
+      "loss": 3.9547,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061879775649698e-05,
+      "loss": 3.9552,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061041180898646e-05,
+      "loss": 3.9531,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060202586147594e-05,
+      "loss": 3.9545,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059363991396542e-05,
+      "loss": 3.9579,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05852539664549e-05,
+      "loss": 3.9515,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057688439774811e-05,
+      "loss": 3.9506,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056849845023759e-05,
+      "loss": 3.9623,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056011250272707e-05,
+      "loss": 3.9555,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055172655521656e-05,
+      "loss": 3.9396,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0543356986509766e-05,
+      "loss": 3.9543,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0534971038999246e-05,
+      "loss": 3.9573,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0526585091488726e-05,
+      "loss": 3.9517,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0518199143978206e-05,
+      "loss": 3.9511,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509813196467686e-05,
+      "loss": 3.9463,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501427248957165e-05,
+      "loss": 3.9516,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0493041301446645e-05,
+      "loss": 3.9487,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048465535393612e-05,
+      "loss": 3.9503,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0476269406425602e-05,
+      "loss": 3.9551,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0467883458915082e-05,
+      "loss": 3.9497,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459513890208298e-05,
+      "loss": 3.9662,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045112794269777e-05,
+      "loss": 3.9368,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.044274199518725e-05,
+      "loss": 3.9446,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.043435604767673e-05,
+      "loss": 3.9512,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.042597010016621e-05,
+      "loss": 3.9497,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0417600531459423e-05,
+      "loss": 3.9598,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0409214583948903e-05,
+      "loss": 3.9521,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0400828636438383e-05,
+      "loss": 3.9452,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392442688927863e-05,
+      "loss": 3.9491,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0384056741417343e-05,
+      "loss": 3.9498,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375703551514288e-05,
+      "loss": 3.9417,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036731760400377e-05,
+      "loss": 3.9422,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358931656493245e-05,
+      "loss": 3.9504,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0350545708982725e-05,
+      "loss": 3.9525,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0342159761472205e-05,
+      "loss": 3.9388,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0333790192765417e-05,
+      "loss": 3.945,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0325404245254897e-05,
+      "loss": 3.9356,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0317018297744377e-05,
+      "loss": 3.9544,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308632350233857e-05,
+      "loss": 3.942,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0300262781527066e-05,
+      "loss": 3.9467,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291876834016546e-05,
+      "loss": 3.9504,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0283490886506026e-05,
+      "loss": 3.956,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.027510493899551e-05,
+      "loss": 3.9489,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026671899148499e-05,
+      "loss": 3.9469,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025833304397447e-05,
+      "loss": 3.9519,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.024994709646395e-05,
+      "loss": 3.9468,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.024156114895343e-05,
+      "loss": 3.9398,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.023317520144291e-05,
+      "loss": 3.9499,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224805632736118e-05,
+      "loss": 3.9475,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.02164196852256e-05,
+      "loss": 3.9547,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.020803373771508e-05,
+      "loss": 3.9417,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019964779020456e-05,
+      "loss": 3.9488,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01912946003015e-05,
+      "loss": 3.9438,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.018290865279098e-05,
+      "loss": 3.9449,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174522705280463e-05,
+      "loss": 3.9373,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0166136757769943e-05,
+      "loss": 3.9552,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0157750810259423e-05,
+      "loss": 3.9616,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0149364862748903e-05,
+      "loss": 3.9455,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0140978915238383e-05,
+      "loss": 3.9421,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0132609346531592e-05,
+      "loss": 3.9545,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0124223399021072e-05,
+      "loss": 3.9502,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0115837451510552e-05,
+      "loss": 3.9449,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0107451504000035e-05,
+      "loss": 3.9503,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0099065556489515e-05,
+      "loss": 3.9591,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090679608978995e-05,
+      "loss": 3.9519,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0082293661468475e-05,
+      "loss": 3.9476,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073907713957955e-05,
+      "loss": 3.9436,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065521766447428e-05,
+      "loss": 3.9521,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.005713581893691e-05,
+      "loss": 3.9515,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0048766250230127e-05,
+      "loss": 3.9376,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0040380302719607e-05,
+      "loss": 3.9619,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.003199435520908e-05,
+      "loss": 3.946,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.002360840769856e-05,
+      "loss": 3.9516,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015238838991776e-05,
+      "loss": 3.9529,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006852891481253e-05,
+      "loss": 3.9436,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.067618370056152,
+      "eval_runtime": 289.2318,
+      "eval_samples_per_second": 1319.326,
+      "eval_steps_per_second": 41.23,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999848332277447e-05,
+      "loss": 3.9454,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999009737526395e-05,
+      "loss": 3.9482,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.998171142775343e-05,
+      "loss": 3.9606,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9973325480242902e-05,
+      "loss": 3.9499,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9964939532732382e-05,
+      "loss": 3.9524,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9956553585221865e-05,
+      "loss": 3.9432,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9948167637711345e-05,
+      "loss": 3.9473,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939781690200825e-05,
+      "loss": 3.9447,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9931412121494034e-05,
+      "loss": 3.9573,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9923026173983514e-05,
+      "loss": 3.9417,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9914640226472994e-05,
+      "loss": 3.9557,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9906270657766207e-05,
+      "loss": 3.9584,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9897884710255687e-05,
+      "loss": 3.9389,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9889498762745166e-05,
+      "loss": 3.9441,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9881112815234646e-05,
+      "loss": 3.9321,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872726867724126e-05,
+      "loss": 3.937,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9864357299017335e-05,
+      "loss": 3.9463,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.985597135150682e-05,
+      "loss": 3.9411,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9847601782800028e-05,
+      "loss": 3.9443,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9839215835289508e-05,
+      "loss": 3.9665,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830829887778988e-05,
+      "loss": 3.9444,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9822443940268468e-05,
+      "loss": 3.9519,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9814057992757948e-05,
+      "loss": 3.9487,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805672045247428e-05,
+      "loss": 3.9441,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.979728609773691e-05,
+      "loss": 3.9474,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.978890015022639e-05,
+      "loss": 3.9343,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.978051420271587e-05,
+      "loss": 3.9439,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.977212825520535e-05,
+      "loss": 3.9447,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.976374230769483e-05,
+      "loss": 3.9327,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.975537273898804e-05,
+      "loss": 3.9452,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.974698679147752e-05,
+      "loss": 3.9397,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9738600843967003e-05,
+      "loss": 3.9517,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9730214896456483e-05,
+      "loss": 3.9463,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9721828948945963e-05,
+      "loss": 3.9376,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9713459380239172e-05,
+      "loss": 3.9427,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9705073432728652e-05,
+      "loss": 3.9608,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9696687485218132e-05,
+      "loss": 3.9363,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9688301537707612e-05,
+      "loss": 3.9434,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967991559019709e-05,
+      "loss": 3.9447,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967152964268657e-05,
+      "loss": 3.9352,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9663160073979785e-05,
+      "loss": 3.944,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9654774126469264e-05,
+      "loss": 3.9564,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.964638817895874e-05,
+      "loss": 3.9526,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963800223144822e-05,
+      "loss": 3.9403,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.96296162839377e-05,
+      "loss": 3.9505,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962123033642718e-05,
+      "loss": 3.9405,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.961284438891666e-05,
+      "loss": 3.9509,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.960445844140614e-05,
+      "loss": 3.9281,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.959608887269935e-05,
+      "loss": 3.9413,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9587719303992562e-05,
+      "loss": 3.951,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9579333356482042e-05,
+      "loss": 3.9497,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9570947408971522e-05,
+      "loss": 3.9434,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9562561461461002e-05,
+      "loss": 3.9325,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9554175513950482e-05,
+      "loss": 3.9452,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9545805945243695e-05,
+      "loss": 3.9313,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9537419997733175e-05,
+      "loss": 3.9571,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9529034050222655e-05,
+      "loss": 3.9374,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9520648102712135e-05,
+      "loss": 3.9573,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9512262155201615e-05,
+      "loss": 3.9418,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9503876207691094e-05,
+      "loss": 3.9344,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9495490260180574e-05,
+      "loss": 3.9458,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9487104312670054e-05,
+      "loss": 3.9379,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9478734743963267e-05,
+      "loss": 3.9387,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9470348796452747e-05,
+      "loss": 3.9354,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9461962848942227e-05,
+      "loss": 3.951,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9453576901431707e-05,
+      "loss": 3.9344,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9445190953921187e-05,
+      "loss": 3.9343,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9436821385214396e-05,
+      "loss": 3.9284,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9428435437703876e-05,
+      "loss": 3.9325,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.942004949019336e-05,
+      "loss": 3.9535,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.941166354268284e-05,
+      "loss": 3.9493,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9403293973976048e-05,
+      "loss": 3.9378,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394908026465528e-05,
+      "loss": 3.9403,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9386522078955008e-05,
+      "loss": 3.9463,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9378136131444488e-05,
+      "loss": 3.947,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.93697665627377e-05,
+      "loss": 3.9401,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.936138061522718e-05,
+      "loss": 3.9469,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.935301104652039e-05,
+      "loss": 3.9413,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.934462509900987e-05,
+      "loss": 3.9492,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.933623915149935e-05,
+      "loss": 3.9401,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327869582792562e-05,
+      "loss": 3.9417,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9319483635282042e-05,
+      "loss": 3.9516,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9311097687771522e-05,
+      "loss": 3.9442,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9302711740261002e-05,
+      "loss": 3.9273,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9294325792750482e-05,
+      "loss": 3.9412,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9285939845239962e-05,
+      "loss": 3.945,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9277553897729442e-05,
+      "loss": 3.9443,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926916795021892e-05,
+      "loss": 3.9407,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9260782002708398e-05,
+      "loss": 3.9349,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9252412434001614e-05,
+      "loss": 3.9426,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9244026486491094e-05,
+      "loss": 3.9357,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235640538980574e-05,
+      "loss": 3.9425,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.922725459147005e-05,
+      "loss": 3.9395,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9218885022763266e-05,
+      "loss": 3.9356,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9210499075252746e-05,
+      "loss": 3.957,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.920211312774222e-05,
+      "loss": 3.9295,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.91937271802317e-05,
+      "loss": 3.9286,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.918534123272118e-05,
+      "loss": 3.9434,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.917695528521066e-05,
+      "loss": 3.9424,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9168569337700143e-05,
+      "loss": 3.9492,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9160183390189623e-05,
+      "loss": 3.9395,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9151813821482832e-05,
+      "loss": 3.9326,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.914344425277604e-05,
+      "loss": 3.9365,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.913505830526552e-05,
+      "loss": 3.9451,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9126672357755004e-05,
+      "loss": 3.9313,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9118286410244484e-05,
+      "loss": 3.929,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9109900462733964e-05,
+      "loss": 3.9427,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9101514515223444e-05,
+      "loss": 3.9399,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9093128567712924e-05,
+      "loss": 3.9296,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9084742620202404e-05,
+      "loss": 3.9313,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9076356672691884e-05,
+      "loss": 3.9299,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9067970725181364e-05,
+      "loss": 3.942,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9059584777670844e-05,
+      "loss": 3.9303,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9051198830160324e-05,
+      "loss": 3.9371,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9042829261453536e-05,
+      "loss": 3.9379,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9034443313943016e-05,
+      "loss": 3.9445,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9026073745236225e-05,
+      "loss": 3.9339,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9017687797725705e-05,
+      "loss": 3.9383,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9009301850215185e-05,
+      "loss": 3.9412,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900091590270467e-05,
+      "loss": 3.9391,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.899252995519415e-05,
+      "loss": 3.9237,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8984160386487358e-05,
+      "loss": 3.9408,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8975774438976838e-05,
+      "loss": 3.9384,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8967388491466318e-05,
+      "loss": 3.9436,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8959002543955798e-05,
+      "loss": 3.9302,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950616596445278e-05,
+      "loss": 3.9415,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.894223064893476e-05,
+      "loss": 3.9322,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.893386108022797e-05,
+      "loss": 3.936,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.892547513271745e-05,
+      "loss": 3.9225,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.891708918520693e-05,
+      "loss": 3.9472,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8908703237696406e-05,
+      "loss": 3.9506,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8900317290185886e-05,
+      "loss": 3.9351,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8891931342675366e-05,
+      "loss": 3.927,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8883545395164846e-05,
+      "loss": 3.9512,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8875159447654326e-05,
+      "loss": 3.9388,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8866773500143806e-05,
+      "loss": 3.9396,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8858403931437015e-05,
+      "loss": 3.9383,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88500179839265e-05,
+      "loss": 3.9473,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8841648415219708e-05,
+      "loss": 3.9381,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8833262467709188e-05,
+      "loss": 3.9372,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8824876520198668e-05,
+      "loss": 3.9306,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8816490572688148e-05,
+      "loss": 3.9419,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8808104625177628e-05,
+      "loss": 3.9413,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879973505647084e-05,
+      "loss": 3.93,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879134910896032e-05,
+      "loss": 3.9521,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.87829631614498e-05,
+      "loss": 3.9321,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.877457721393928e-05,
+      "loss": 3.9413,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.876619126642876e-05,
+      "loss": 3.9421,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.875780531891824e-05,
+      "loss": 3.9344,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.061960697174072,
+      "eval_runtime": 288.7843,
+      "eval_samples_per_second": 1321.37,
+      "eval_steps_per_second": 41.294,
+      "step": 1297440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.874941937140772e-05,
+      "loss": 3.9308,
+      "step": 1297920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.87410334238972e-05,
+      "loss": 3.9406,
+      "step": 1298432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8732647476386683e-05,
+      "loss": 3.948,
+      "step": 1298944
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.872429428648362e-05,
+      "loss": 3.9404,
+      "step": 1299456
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.87159083389731e-05,
+      "loss": 3.945,
+      "step": 1299968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.870752239146258e-05,
+      "loss": 3.9314,
+      "step": 1300480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8699152822755794e-05,
+      "loss": 3.9333,
+      "step": 1300992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8690766875245274e-05,
+      "loss": 3.9331,
+      "step": 1301504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8682380927734754e-05,
+      "loss": 3.9475,
+      "step": 1302016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8673994980224234e-05,
+      "loss": 3.9344,
+      "step": 1302528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8665609032713714e-05,
+      "loss": 3.9437,
+      "step": 1303040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8657223085203193e-05,
+      "loss": 3.9451,
+      "step": 1303552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8648853516496406e-05,
+      "loss": 3.9311,
+      "step": 1304064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8640467568985886e-05,
+      "loss": 3.9323,
+      "step": 1304576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8632081621475366e-05,
+      "loss": 3.9216,
+      "step": 1305088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8623695673964846e-05,
+      "loss": 3.9305,
+      "step": 1305600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8615309726454326e-05,
+      "loss": 3.9307,
+      "step": 1306112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8606923778943806e-05,
+      "loss": 3.9319,
+      "step": 1306624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8598537831433286e-05,
+      "loss": 3.93,
+      "step": 1307136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8590151883922766e-05,
+      "loss": 3.9574,
+      "step": 1307648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8581765936412242e-05,
+      "loss": 3.9323,
+      "step": 1308160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8573379988901722e-05,
+      "loss": 3.9368,
+      "step": 1308672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8564994041391202e-05,
+      "loss": 3.9422,
+      "step": 1309184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8556624472684418e-05,
+      "loss": 3.9364,
+      "step": 1309696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.854823852517389e-05,
+      "loss": 3.9319,
+      "step": 1310208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8539852577663375e-05,
+      "loss": 3.9265,
+      "step": 1310720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.853148300895659e-05,
+      "loss": 3.933,
+      "step": 1311232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8523097061446064e-05,
+      "loss": 3.9322,
+      "step": 1311744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8514711113935544e-05,
+      "loss": 3.9215,
+      "step": 1312256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8506325166425024e-05,
+      "loss": 3.9364,
+      "step": 1312768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8497939218914503e-05,
+      "loss": 3.9259,
+      "step": 1313280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8489569650207716e-05,
+      "loss": 3.9415,
+      "step": 1313792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8481183702697196e-05,
+      "loss": 3.9405,
+      "step": 1314304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8472797755186676e-05,
+      "loss": 3.9242,
+      "step": 1314816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8464411807676156e-05,
+      "loss": 3.9372,
+      "step": 1315328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8456025860165636e-05,
+      "loss": 3.9475,
+      "step": 1315840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8447639912655116e-05,
+      "loss": 3.9241,
+      "step": 1316352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8439253965144596e-05,
+      "loss": 3.9367,
+      "step": 1316864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8430868017634076e-05,
+      "loss": 3.9299,
+      "step": 1317376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8422498448927288e-05,
+      "loss": 3.9254,
+      "step": 1317888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8414128880220497e-05,
+      "loss": 3.9338,
+      "step": 1318400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8405742932709977e-05,
+      "loss": 3.9443,
+      "step": 1318912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8397356985199457e-05,
+      "loss": 3.9408,
+      "step": 1319424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8388971037688937e-05,
+      "loss": 3.9302,
+      "step": 1319936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8380585090178417e-05,
+      "loss": 3.941,
+      "step": 1320448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.83721991426679e-05,
+      "loss": 3.928,
+      "step": 1320960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.836381319515738e-05,
+      "loss": 3.9377,
+      "step": 1321472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.835542724764686e-05,
+      "loss": 3.9193,
+      "step": 1321984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.834705767894007e-05,
+      "loss": 3.9269,
+      "step": 1322496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.833867173142955e-05,
+      "loss": 3.9404,
+      "step": 1323008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.833028578391903e-05,
+      "loss": 3.9419,
+      "step": 1323520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.832189983640851e-05,
+      "loss": 3.9301,
+      "step": 1324032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8313530267701722e-05,
+      "loss": 3.9242,
+      "step": 1324544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8305144320191202e-05,
+      "loss": 3.9364,
+      "step": 1325056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.829675837268068e-05,
+      "loss": 3.9176,
+      "step": 1325568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.828838880397389e-05,
+      "loss": 3.9438,
+      "step": 1326080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.828000285646337e-05,
+      "loss": 3.9296,
+      "step": 1326592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8271616908952854e-05,
+      "loss": 3.9462,
+      "step": 1327104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8263230961442334e-05,
+      "loss": 3.9324,
+      "step": 1327616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8254861392735543e-05,
+      "loss": 3.9249,
+      "step": 1328128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8246475445225023e-05,
+      "loss": 3.9298,
+      "step": 1328640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8238089497714503e-05,
+      "loss": 3.9346,
+      "step": 1329152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8229703550203983e-05,
+      "loss": 3.9235,
+      "step": 1329664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8221317602693463e-05,
+      "loss": 3.9261,
+      "step": 1330176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8212931655182946e-05,
+      "loss": 3.9414,
+      "step": 1330688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8204545707672426e-05,
+      "loss": 3.928,
+      "step": 1331200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8196176138965635e-05,
+      "loss": 3.92,
+      "step": 1331712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8187806570258844e-05,
+      "loss": 3.9224,
+      "step": 1332224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8179420622748324e-05,
+      "loss": 3.9229,
+      "step": 1332736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8171034675237808e-05,
+      "loss": 3.9432,
+      "step": 1333248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8162648727727288e-05,
+      "loss": 3.939,
+      "step": 1333760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8154262780216768e-05,
+      "loss": 3.9279,
+      "step": 1334272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8145876832706248e-05,
+      "loss": 3.9314,
+      "step": 1334784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8137490885195728e-05,
+      "loss": 3.9368,
+      "step": 1335296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.81291049376852e-05,
+      "loss": 3.935,
+      "step": 1335808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8120718990174684e-05,
+      "loss": 3.9287,
+      "step": 1336320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8112333042664164e-05,
+      "loss": 3.9375,
+      "step": 1336832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8103963473957373e-05,
+      "loss": 3.9312,
+      "step": 1337344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8095577526446853e-05,
+      "loss": 3.9405,
+      "step": 1337856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8087191578936333e-05,
+      "loss": 3.9285,
+      "step": 1338368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8078805631425813e-05,
+      "loss": 3.9325,
+      "step": 1338880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8070419683915293e-05,
+      "loss": 3.937,
+      "step": 1339392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8062033736404776e-05,
+      "loss": 3.9366,
+      "step": 1339904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8053647788894256e-05,
+      "loss": 3.9185,
+      "step": 1340416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8045261841383736e-05,
+      "loss": 3.9314,
+      "step": 1340928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8036892272676945e-05,
+      "loss": 3.9324,
+      "step": 1341440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8028506325166425e-05,
+      "loss": 3.9351,
+      "step": 1341952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8020120377655905e-05,
+      "loss": 3.9307,
+      "step": 1342464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8011750808949118e-05,
+      "loss": 3.9236,
+      "step": 1342976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.8003364861438598e-05,
+      "loss": 3.9319,
+      "step": 1343488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7994978913928078e-05,
+      "loss": 3.9256,
+      "step": 1344000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7986592966417558e-05,
+      "loss": 3.935,
+      "step": 1344512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7978207018907038e-05,
+      "loss": 3.9239,
+      "step": 1345024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7969821071396517e-05,
+      "loss": 3.9302,
+      "step": 1345536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.796145150268973e-05,
+      "loss": 3.9408,
+      "step": 1346048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.795306555517921e-05,
+      "loss": 3.9327,
+      "step": 1346560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.794467960766869e-05,
+      "loss": 3.9124,
+      "step": 1347072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.793629366015817e-05,
+      "loss": 3.932,
+      "step": 1347584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.792790771264765e-05,
+      "loss": 3.9332,
+      "step": 1348096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.791952176513713e-05,
+      "loss": 3.9367,
+      "step": 1348608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.791113581762661e-05,
+      "loss": 3.928,
+      "step": 1349120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7902766248919822e-05,
+      "loss": 3.927,
+      "step": 1349632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7894380301409302e-05,
+      "loss": 3.9248,
+      "step": 1350144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7885994353898782e-05,
+      "loss": 3.9347,
+      "step": 1350656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7877608406388262e-05,
+      "loss": 3.9176,
+      "step": 1351168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7869222458877735e-05,
+      "loss": 3.9248,
+      "step": 1351680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7860836511367215e-05,
+      "loss": 3.9309,
+      "step": 1352192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.785246694266043e-05,
+      "loss": 3.9315,
+      "step": 1352704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.784408099514991e-05,
+      "loss": 3.9118,
+      "step": 1353216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7835695047639388e-05,
+      "loss": 3.9271,
+      "step": 1353728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7827309100128868e-05,
+      "loss": 3.9157,
+      "step": 1354240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7818923152618347e-05,
+      "loss": 3.9324,
+      "step": 1354752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7810537205107827e-05,
+      "loss": 3.9218,
+      "step": 1355264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7802151257597307e-05,
+      "loss": 3.9182,
+      "step": 1355776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7793765310086787e-05,
+      "loss": 3.9362,
+      "step": 1356288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.778539574138e-05,
+      "loss": 3.9334,
+      "step": 1356800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.777702617267321e-05,
+      "loss": 3.9254,
+      "step": 1357312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.776864022516269e-05,
+      "loss": 3.9304,
+      "step": 1357824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.776025427765217e-05,
+      "loss": 3.9255,
+      "step": 1358336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.775186833014165e-05,
+      "loss": 3.9321,
+      "step": 1358848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7743482382631132e-05,
+      "loss": 3.9126,
+      "step": 1359360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7735096435120612e-05,
+      "loss": 3.9359,
+      "step": 1359872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7726710487610092e-05,
+      "loss": 3.9239,
+      "step": 1360384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.77183409189033e-05,
+      "loss": 3.931,
+      "step": 1360896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.770995497139278e-05,
+      "loss": 3.9218,
+      "step": 1361408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.770156902388226e-05,
+      "loss": 3.9296,
+      "step": 1361920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.769318307637174e-05,
+      "loss": 3.9225,
+      "step": 1362432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7684813507664953e-05,
+      "loss": 3.927,
+      "step": 1362944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7676427560154433e-05,
+      "loss": 3.9113,
+      "step": 1363456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7668041612643913e-05,
+      "loss": 3.9401,
+      "step": 1363968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7659655665133393e-05,
+      "loss": 3.9353,
+      "step": 1364480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7651269717622873e-05,
+      "loss": 3.9268,
+      "step": 1364992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7642883770112353e-05,
+      "loss": 3.9192,
+      "step": 1365504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7634514201405566e-05,
+      "loss": 3.9355,
+      "step": 1366016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7626128253895046e-05,
+      "loss": 3.934,
+      "step": 1366528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7617742306384526e-05,
+      "loss": 3.925,
+      "step": 1367040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7609356358874006e-05,
+      "loss": 3.9223,
+      "step": 1367552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7600970411363486e-05,
+      "loss": 3.9371,
+      "step": 1368064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7592584463852966e-05,
+      "loss": 3.927,
+      "step": 1368576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7584214895146178e-05,
+      "loss": 3.9358,
+      "step": 1369088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7575828947635658e-05,
+      "loss": 3.9183,
+      "step": 1369600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7567443000125138e-05,
+      "loss": 3.9309,
+      "step": 1370112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7559057052614618e-05,
+      "loss": 3.9298,
+      "step": 1370624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7550671105104098e-05,
+      "loss": 3.9207,
+      "step": 1371136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.754228515759357e-05,
+      "loss": 3.9389,
+      "step": 1371648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7533899210083054e-05,
+      "loss": 3.9228,
+      "step": 1372160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7525513262572534e-05,
+      "loss": 3.9296,
+      "step": 1372672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.751714369386575e-05,
+      "loss": 3.9278,
+      "step": 1373184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.750877412515896e-05,
+      "loss": 3.9263,
+      "step": 1373696
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.057029724121094,
+      "eval_runtime": 289.3607,
+      "eval_samples_per_second": 1318.738,
+      "eval_steps_per_second": 41.212,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.750038817764844e-05,
+      "loss": 3.9207,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.749200223013792e-05,
+      "loss": 3.9307,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483616282627396e-05,
+      "loss": 3.9363,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7475230335116876e-05,
+      "loss": 3.9314,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466844387606356e-05,
+      "loss": 3.9365,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458458440095836e-05,
+      "loss": 3.92,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7450072492585316e-05,
+      "loss": 3.9232,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7441686545074796e-05,
+      "loss": 3.924,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7433300597564275e-05,
+      "loss": 3.9355,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7424914650053755e-05,
+      "loss": 3.9244,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741652870254324e-05,
+      "loss": 3.9343,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740814275503272e-05,
+      "loss": 3.9353,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7399773186325928e-05,
+      "loss": 3.9199,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391387238815408e-05,
+      "loss": 3.9237,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7383001291304888e-05,
+      "loss": 3.9094,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374615343794368e-05,
+      "loss": 3.9214,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7366229396283848e-05,
+      "loss": 3.923,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357843448773328e-05,
+      "loss": 3.9206,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734945750126281e-05,
+      "loss": 3.9204,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7341071553752284e-05,
+      "loss": 3.9454,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332685606241764e-05,
+      "loss": 3.9253,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324299658731244e-05,
+      "loss": 3.9255,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7315913711220724e-05,
+      "loss": 3.9327,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7307527763710204e-05,
+      "loss": 3.9284,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7299158195003416e-05,
+      "loss": 3.9204,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290772247492896e-05,
+      "loss": 3.9166,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282386299982376e-05,
+      "loss": 3.9268,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7274000352471856e-05,
+      "loss": 3.9184,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265630783765065e-05,
+      "loss": 3.9149,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.725724483625455e-05,
+      "loss": 3.9274,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724885888874403e-05,
+      "loss": 3.9194,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724047294123351e-05,
+      "loss": 3.931,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.723208699372299e-05,
+      "loss": 3.9278,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223717425016198e-05,
+      "loss": 3.9152,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.721534785630941e-05,
+      "loss": 3.9277,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720696190879889e-05,
+      "loss": 3.9398,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719857596128837e-05,
+      "loss": 3.9107,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719019001377785e-05,
+      "loss": 3.9294,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718180406626733e-05,
+      "loss": 3.9175,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.717341811875681e-05,
+      "loss": 3.9211,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.716504855005002e-05,
+      "loss": 3.9214,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156662602539502e-05,
+      "loss": 3.9374,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7148276655028982e-05,
+      "loss": 3.93,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139890707518462e-05,
+      "loss": 3.9228,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131504760007942e-05,
+      "loss": 3.9284,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7123118812497422e-05,
+      "loss": 3.9183,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114732864986902e-05,
+      "loss": 3.9315,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106346917476382e-05,
+      "loss": 3.9121,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097977348769595e-05,
+      "loss": 3.9175,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089591401259075e-05,
+      "loss": 3.9287,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7081205453748555e-05,
+      "loss": 3.9281,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072819506238035e-05,
+      "loss": 3.9214,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064433558727514e-05,
+      "loss": 3.9173,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7056063990020724e-05,
+      "loss": 3.9264,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7047678042510204e-05,
+      "loss": 3.9088,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7039308473803416e-05,
+      "loss": 3.9301,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030922526292896e-05,
+      "loss": 3.9232,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022536578782376e-05,
+      "loss": 3.9349,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7014150631271856e-05,
+      "loss": 3.9278,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7005781062565065e-05,
+      "loss": 3.9179,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.699739511505455e-05,
+      "loss": 3.9197,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6989009167544028e-05,
+      "loss": 3.9262,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6980623220033508e-05,
+      "loss": 3.9123,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6972237272522988e-05,
+      "loss": 3.9165,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6963851325012468e-05,
+      "loss": 3.9274,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.695546537750194e-05,
+      "loss": 3.9221,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6947079429991425e-05,
+      "loss": 3.9091,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693870986128464e-05,
+      "loss": 3.9129,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693032391377412e-05,
+      "loss": 3.9118,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.692195434506733e-05,
+      "loss": 3.9339,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.691356839755681e-05,
+      "loss": 3.9276,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.690518245004629e-05,
+      "loss": 3.9182,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689679650253577e-05,
+      "loss": 3.9239,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888410555025246e-05,
+      "loss": 3.9219,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6880024607514726e-05,
+      "loss": 3.929,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871638660004206e-05,
+      "loss": 3.9166,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.686328547010115e-05,
+      "loss": 3.929,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.685489952259063e-05,
+      "loss": 3.9257,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.684651357508011e-05,
+      "loss": 3.9272,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6838127627569594e-05,
+      "loss": 3.9182,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6829741680059067e-05,
+      "loss": 3.9235,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821355732548547e-05,
+      "loss": 3.9257,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812969785038027e-05,
+      "loss": 3.9275,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804583837527507e-05,
+      "loss": 3.914,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6796197890016987e-05,
+      "loss": 3.9215,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.678781194250647e-05,
+      "loss": 3.9213,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.677944237379968e-05,
+      "loss": 3.9251,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.677105642628916e-05,
+      "loss": 3.9278,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.676267047877864e-05,
+      "loss": 3.9107,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.675430091007185e-05,
+      "loss": 3.9213,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745914962561332e-05,
+      "loss": 3.9198,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737529015050812e-05,
+      "loss": 3.9286,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6729143067540292e-05,
+      "loss": 3.9101,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720757120029772e-05,
+      "loss": 3.9223,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712371172519252e-05,
+      "loss": 3.9327,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703985225008732e-05,
+      "loss": 3.9217,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695599277498212e-05,
+      "loss": 3.9042,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.668722970879142e-05,
+      "loss": 3.9203,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678843761280904e-05,
+      "loss": 3.9252,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670457813770384e-05,
+      "loss": 3.9253,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6662071866259864e-05,
+      "loss": 3.9204,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653685918749344e-05,
+      "loss": 3.9149,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645299971238824e-05,
+      "loss": 3.9162,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636930402532033e-05,
+      "loss": 3.9233,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628544455021513e-05,
+      "loss": 3.9081,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6620158507510996e-05,
+      "loss": 3.9161,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6611788938804205e-05,
+      "loss": 3.9209,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6603402991293685e-05,
+      "loss": 3.923,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6595017043783165e-05,
+      "loss": 3.9016,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6586631096272645e-05,
+      "loss": 3.92,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6578245148762125e-05,
+      "loss": 3.9074,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6569859201251602e-05,
+      "loss": 3.9219,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6561473253741082e-05,
+      "loss": 3.9161,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6553087306230562e-05,
+      "loss": 3.9086,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544717737523778e-05,
+      "loss": 3.9256,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.653633179001325e-05,
+      "loss": 3.9255,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527945842502734e-05,
+      "loss": 3.9131,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519559894992214e-05,
+      "loss": 3.9193,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6511173947481694e-05,
+      "loss": 3.9193,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502787999971174e-05,
+      "loss": 3.9213,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494418431264383e-05,
+      "loss": 3.9065,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6486032483753863e-05,
+      "loss": 3.9239,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477646536243343e-05,
+      "loss": 3.9171,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6469260588732826e-05,
+      "loss": 3.9209,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460874641222306e-05,
+      "loss": 3.9164,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452505072515515e-05,
+      "loss": 3.9239,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6444119125004995e-05,
+      "loss": 3.9057,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6435733177494475e-05,
+      "loss": 3.9169,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6427347229983955e-05,
+      "loss": 3.9034,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6418961282473435e-05,
+      "loss": 3.9315,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6410591713766648e-05,
+      "loss": 3.9226,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6402205766256128e-05,
+      "loss": 3.9217,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6393819818745608e-05,
+      "loss": 3.9101,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385433871235088e-05,
+      "loss": 3.9277,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6377047923724568e-05,
+      "loss": 3.927,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368661976214048e-05,
+      "loss": 3.914,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360276028703527e-05,
+      "loss": 3.9173,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.635189008119301e-05,
+      "loss": 3.9247,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.634352051248622e-05,
+      "loss": 3.918,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.63351345649757e-05,
+      "loss": 3.9278,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.632676499626891e-05,
+      "loss": 3.9101,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.631837904875839e-05,
+      "loss": 3.9235,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309993101247872e-05,
+      "loss": 3.9173,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301607153737352e-05,
+      "loss": 3.9097,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6293221206226832e-05,
+      "loss": 3.9316,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.628485163752004e-05,
+      "loss": 3.9133,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.627646569000952e-05,
+      "loss": 3.9192,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6268096121302734e-05,
+      "loss": 3.9209,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259710173792214e-05,
+      "loss": 3.9151,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0537848472595215,
+      "eval_runtime": 287.4232,
+      "eval_samples_per_second": 1327.628,
+      "eval_steps_per_second": 41.489,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6251324226281694e-05,
+      "loss": 3.9171,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242938278771174e-05,
+      "loss": 3.9198,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234552331260654e-05,
+      "loss": 3.927,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6226166383750134e-05,
+      "loss": 3.9217,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217780436239613e-05,
+      "loss": 3.928,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.620939448872909e-05,
+      "loss": 3.914,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6201024920022306e-05,
+      "loss": 3.9161,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192655351315515e-05,
+      "loss": 3.9071,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6184269403804995e-05,
+      "loss": 3.9319,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6175883456294475e-05,
+      "loss": 3.9145,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167497508783955e-05,
+      "loss": 3.9249,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6159111561273435e-05,
+      "loss": 3.9245,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6150741992566647e-05,
+      "loss": 3.9181,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6142356045056127e-05,
+      "loss": 3.9087,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6133970097545607e-05,
+      "loss": 3.9082,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6125584150035087e-05,
+      "loss": 3.9053,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6117198202524564e-05,
+      "loss": 3.9168,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6108812255014044e-05,
+      "loss": 3.9115,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6100426307503524e-05,
+      "loss": 3.9127,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6092040359993004e-05,
+      "loss": 3.9352,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083654412482484e-05,
+      "loss": 3.9183,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6075268464971964e-05,
+      "loss": 3.9184,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6066882517461443e-05,
+      "loss": 3.9227,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6058496569950923e-05,
+      "loss": 3.9199,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6050127001244136e-05,
+      "loss": 3.9137,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041741053733616e-05,
+      "loss": 3.9053,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033355106223096e-05,
+      "loss": 3.9214,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6024985537516305e-05,
+      "loss": 3.9047,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6016599590005785e-05,
+      "loss": 3.9099,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6008213642495265e-05,
+      "loss": 3.9131,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5999827694984745e-05,
+      "loss": 3.9099,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5991458126277957e-05,
+      "loss": 3.9222,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5983072178767437e-05,
+      "loss": 3.9173,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5974686231256917e-05,
+      "loss": 3.9089,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966300283746397e-05,
+      "loss": 3.9189,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957914336235877e-05,
+      "loss": 3.9269,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.594954476752909e-05,
+      "loss": 3.9064,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.594115882001857e-05,
+      "loss": 3.9203,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.593277287250805e-05,
+      "loss": 3.9043,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.592438692499753e-05,
+      "loss": 3.918,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.591601735629074e-05,
+      "loss": 3.9086,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.590763140878022e-05,
+      "loss": 3.9298,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.58992454612697e-05,
+      "loss": 3.9173,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890859513759182e-05,
+      "loss": 3.9145,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5882473566248662e-05,
+      "loss": 3.9213,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5874087618738142e-05,
+      "loss": 3.9111,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.586570167122762e-05,
+      "loss": 3.9209,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5857315723717095e-05,
+      "loss": 3.9073,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584894615501031e-05,
+      "loss": 3.905,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5840576586303523e-05,
+      "loss": 3.9215,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5832190638793003e-05,
+      "loss": 3.9173,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5823821070086212e-05,
+      "loss": 3.9177,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815435122575692e-05,
+      "loss": 3.9063,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5807049175065172e-05,
+      "loss": 3.9158,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798663227554652e-05,
+      "loss": 3.9008,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5790277280044135e-05,
+      "loss": 3.92,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781891332533615e-05,
+      "loss": 3.914,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773505385023095e-05,
+      "loss": 3.926,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.576511943751257e-05,
+      "loss": 3.9189,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.575673349000205e-05,
+      "loss": 3.9102,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.574834754249153e-05,
+      "loss": 3.9078,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739961594981012e-05,
+      "loss": 3.9171,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731575647470492e-05,
+      "loss": 3.9051,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.57232060787637e-05,
+      "loss": 3.9087,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.571482013125318e-05,
+      "loss": 3.9158,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.570643418374266e-05,
+      "loss": 3.9173,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569804823623214e-05,
+      "loss": 3.8983,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568966228872162e-05,
+      "loss": 3.9042,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5681292720014833e-05,
+      "loss": 3.9013,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5672906772504313e-05,
+      "loss": 3.9271,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664520824993793e-05,
+      "loss": 3.9173,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5656151256287002e-05,
+      "loss": 3.911,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5647765308776482e-05,
+      "loss": 3.9153,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5639379361265965e-05,
+      "loss": 3.9107,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630993413755445e-05,
+      "loss": 3.9251,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622607466244925e-05,
+      "loss": 3.9028,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.561425427634187e-05,
+      "loss": 3.9261,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5605868328831344e-05,
+      "loss": 3.9111,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5597482381320827e-05,
+      "loss": 3.9186,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5589096433810307e-05,
+      "loss": 3.9135,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5580710486299787e-05,
+      "loss": 3.9121,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5572324538789267e-05,
+      "loss": 3.9224,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5563938591278747e-05,
+      "loss": 3.9151,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5555552643768227e-05,
+      "loss": 3.9056,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5547166696257707e-05,
+      "loss": 3.9146,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553879712755092e-05,
+      "loss": 3.9078,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.55304111800404e-05,
+      "loss": 3.919,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.552202523252988e-05,
+      "loss": 3.9162,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.551363928501936e-05,
+      "loss": 3.9016,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.550525333750884e-05,
+      "loss": 3.9161,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.549686738999832e-05,
+      "loss": 3.9068,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54884814424878e-05,
+      "loss": 3.9159,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.548009549497728e-05,
+      "loss": 3.9046,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5471709547466755e-05,
+      "loss": 3.9143,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.546333997875997e-05,
+      "loss": 3.9213,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.545495403124945e-05,
+      "loss": 3.9102,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.544656808373893e-05,
+      "loss": 3.9013,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5438182136228404e-05,
+      "loss": 3.9109,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542981256752162e-05,
+      "loss": 3.9102,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54214266200111e-05,
+      "loss": 3.9242,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5413040672500584e-05,
+      "loss": 3.9118,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5404654724990057e-05,
+      "loss": 3.9065,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5396268777479537e-05,
+      "loss": 3.9082,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5387899208772753e-05,
+      "loss": 3.9173,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537951326126223e-05,
+      "loss": 3.8981,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537112731375171e-05,
+      "loss": 3.9096,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.536274136624119e-05,
+      "loss": 3.9087,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.535435541873067e-05,
+      "loss": 3.9166,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.534596947122015e-05,
+      "loss": 3.8917,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.533758352370963e-05,
+      "loss": 3.9109,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532919757619911e-05,
+      "loss": 3.9064,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532082800749232e-05,
+      "loss": 3.9044,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.53124420599818e-05,
+      "loss": 3.9119,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.530405611247128e-05,
+      "loss": 3.8995,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.529567016496076e-05,
+      "loss": 3.9182,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.528730059625397e-05,
+      "loss": 3.9152,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278931027547183e-05,
+      "loss": 3.9053,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5270545080036663e-05,
+      "loss": 3.9128,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5262159132526143e-05,
+      "loss": 3.9075,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253773185015623e-05,
+      "loss": 3.9147,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245387237505103e-05,
+      "loss": 3.8937,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5237001289994583e-05,
+      "loss": 3.919,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228615342484063e-05,
+      "loss": 3.9047,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5220229394973542e-05,
+      "loss": 3.9109,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211859826266755e-05,
+      "loss": 3.9087,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5203473878756235e-05,
+      "loss": 3.9174,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5195104310049444e-05,
+      "loss": 3.9032,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5186718362538924e-05,
+      "loss": 3.9034,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5178332415028404e-05,
+      "loss": 3.8989,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5169946467517884e-05,
+      "loss": 3.9184,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5161560520007367e-05,
+      "loss": 3.9152,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5153174572496847e-05,
+      "loss": 3.9122,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5144788624986327e-05,
+      "loss": 3.9021,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5136402677475807e-05,
+      "loss": 3.9184,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5128016729965287e-05,
+      "loss": 3.9168,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5119647161258496e-05,
+      "loss": 3.9063,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511127759255171e-05,
+      "loss": 3.9068,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.510289164504119e-05,
+      "loss": 3.9174,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.509450569753067e-05,
+      "loss": 3.9152,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.508611975002015e-05,
+      "loss": 3.917,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.507773380250963e-05,
+      "loss": 3.9036,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506934785499911e-05,
+      "loss": 3.9132,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506096190748859e-05,
+      "loss": 3.9082,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.50525923387818e-05,
+      "loss": 3.9054,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.504420639127128e-05,
+      "loss": 3.9186,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.503582044376076e-05,
+      "loss": 3.9058,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.502743449625024e-05,
+      "loss": 3.9136,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5019048548739714e-05,
+      "loss": 3.9085,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.501067898003293e-05,
+      "loss": 3.9114,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.049144268035889,
+      "eval_runtime": 287.5835,
+      "eval_samples_per_second": 1326.888,
+      "eval_steps_per_second": 41.466,
+      "step": 1526400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.500229303252241e-05,
+      "loss": 3.9089,
+      "step": 1526784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.499390708501189e-05,
+      "loss": 3.9132,
+      "step": 1527296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.498552113750137e-05,
+      "loss": 3.9175,
+      "step": 1527808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.497713518999085e-05,
+      "loss": 3.9144,
+      "step": 1528320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.496874924248033e-05,
+      "loss": 3.921,
+      "step": 1528832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4960363294969806e-05,
+      "loss": 3.9025,
+      "step": 1529344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4951993726263022e-05,
+      "loss": 3.907,
+      "step": 1529856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4943607778752502e-05,
+      "loss": 3.8997,
+      "step": 1530368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.493522183124198e-05,
+      "loss": 3.9183,
+      "step": 1530880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.492683588373146e-05,
+      "loss": 3.9116,
+      "step": 1531392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.491844993622094e-05,
+      "loss": 3.9146,
+      "step": 1531904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.491006398871042e-05,
+      "loss": 3.9129,
+      "step": 1532416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.490169442000363e-05,
+      "loss": 3.9093,
+      "step": 1532928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.489330847249311e-05,
+      "loss": 3.9022,
+      "step": 1533440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.488492252498259e-05,
+      "loss": 3.9047,
+      "step": 1533952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.487653657747207e-05,
+      "loss": 3.8946,
+      "step": 1534464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.486815062996155e-05,
+      "loss": 3.9088,
+      "step": 1534976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.485976468245103e-05,
+      "loss": 3.9026,
+      "step": 1535488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.485139511374424e-05,
+      "loss": 3.9002,
+      "step": 1536000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4843009166233723e-05,
+      "loss": 3.932,
+      "step": 1536512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4834623218723203e-05,
+      "loss": 3.9037,
+      "step": 1537024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4826237271212683e-05,
+      "loss": 3.9129,
+      "step": 1537536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.481785132370216e-05,
+      "loss": 3.9147,
+      "step": 1538048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4809481754995372e-05,
+      "loss": 3.9106,
+      "step": 1538560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4801095807484852e-05,
+      "loss": 3.9067,
+      "step": 1539072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4792709859974332e-05,
+      "loss": 3.8988,
+      "step": 1539584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4784323912463812e-05,
+      "loss": 3.9131,
+      "step": 1540096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4775954343757024e-05,
+      "loss": 3.8927,
+      "step": 1540608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4767568396246504e-05,
+      "loss": 3.9033,
+      "step": 1541120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4759182448735984e-05,
+      "loss": 3.9028,
+      "step": 1541632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.475079650122546e-05,
+      "loss": 3.9023,
+      "step": 1542144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4742426932518677e-05,
+      "loss": 3.915,
+      "step": 1542656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4734040985008157e-05,
+      "loss": 3.9106,
+      "step": 1543168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4725655037497633e-05,
+      "loss": 3.905,
+      "step": 1543680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4717269089987113e-05,
+      "loss": 3.91,
+      "step": 1544192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4708883142476593e-05,
+      "loss": 3.9166,
+      "step": 1544704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4700497194966073e-05,
+      "loss": 3.9018,
+      "step": 1545216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4692111247455553e-05,
+      "loss": 3.9113,
+      "step": 1545728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4683725299945033e-05,
+      "loss": 3.897,
+      "step": 1546240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4675355731238246e-05,
+      "loss": 3.9102,
+      "step": 1546752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4666969783727725e-05,
+      "loss": 3.8968,
+      "step": 1547264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4658583836217205e-05,
+      "loss": 3.9243,
+      "step": 1547776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4650197888706685e-05,
+      "loss": 3.91,
+      "step": 1548288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4641811941196165e-05,
+      "loss": 3.9061,
+      "step": 1548800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4633425993685645e-05,
+      "loss": 3.9128,
+      "step": 1549312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4625040046175125e-05,
+      "loss": 3.9031,
+      "step": 1549824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4616654098664605e-05,
+      "loss": 3.9103,
+      "step": 1550336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4608284529957814e-05,
+      "loss": 3.9003,
+      "step": 1550848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4599914961251027e-05,
+      "loss": 3.8968,
+      "step": 1551360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4591529013740507e-05,
+      "loss": 3.9169,
+      "step": 1551872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4583143066229987e-05,
+      "loss": 3.9066,
+      "step": 1552384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4574757118719467e-05,
+      "loss": 3.909,
+      "step": 1552896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4566371171208947e-05,
+      "loss": 3.8994,
+      "step": 1553408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4557985223698427e-05,
+      "loss": 3.9053,
+      "step": 1553920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4549599276187907e-05,
+      "loss": 3.8968,
+      "step": 1554432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4541229707481116e-05,
+      "loss": 3.9085,
+      "step": 1554944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.45328437599706e-05,
+      "loss": 3.9044,
+      "step": 1555456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.452445781246008e-05,
+      "loss": 3.9205,
+      "step": 1555968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.451607186494956e-05,
+      "loss": 3.9101,
+      "step": 1556480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.450768591743904e-05,
+      "loss": 3.901,
+      "step": 1556992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4499316348732248e-05,
+      "loss": 3.9022,
+      "step": 1557504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4490930401221728e-05,
+      "loss": 3.9104,
+      "step": 1558016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.448256083251494e-05,
+      "loss": 3.8956,
+      "step": 1558528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.447417488500442e-05,
+      "loss": 3.9037,
+      "step": 1559040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.44657889374939e-05,
+      "loss": 3.9059,
+      "step": 1559552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.445740298998338e-05,
+      "loss": 3.9116,
+      "step": 1560064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.444901704247286e-05,
+      "loss": 3.8895,
+      "step": 1560576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.444063109496234e-05,
+      "loss": 3.8989,
+      "step": 1561088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.443224514745182e-05,
+      "loss": 3.8939,
+      "step": 1561600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.44238591999413e-05,
+      "loss": 3.9157,
+      "step": 1562112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.441547325243078e-05,
+      "loss": 3.9065,
+      "step": 1562624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.440708730492026e-05,
+      "loss": 3.9054,
+      "step": 1563136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.439870135740974e-05,
+      "loss": 3.9067,
+      "step": 1563648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.439031540989922e-05,
+      "loss": 3.9027,
+      "step": 1564160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.438194584119243e-05,
+      "loss": 3.9227,
+      "step": 1564672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.437357627248564e-05,
+      "loss": 3.893,
+      "step": 1565184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.436519032497512e-05,
+      "loss": 3.9131,
+      "step": 1565696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.43568043774646e-05,
+      "loss": 3.909,
+      "step": 1566208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.434841842995408e-05,
+      "loss": 3.9086,
+      "step": 1566720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.434003248244356e-05,
+      "loss": 3.9063,
+      "step": 1567232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.433164653493304e-05,
+      "loss": 3.9044,
+      "step": 1567744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4323276966226254e-05,
+      "loss": 3.9112,
+      "step": 1568256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4314891018715734e-05,
+      "loss": 3.9092,
+      "step": 1568768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4306505071205214e-05,
+      "loss": 3.8956,
+      "step": 1569280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4298119123694694e-05,
+      "loss": 3.9086,
+      "step": 1569792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4289733176184174e-05,
+      "loss": 3.8993,
+      "step": 1570304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.428134722867365e-05,
+      "loss": 3.9112,
+      "step": 1570816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.427296128116313e-05,
+      "loss": 3.9069,
+      "step": 1571328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4264591712456346e-05,
+      "loss": 3.8942,
+      "step": 1571840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4256205764945826e-05,
+      "loss": 3.9069,
+      "step": 1572352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4247819817435302e-05,
+      "loss": 3.901,
+      "step": 1572864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4239433869924782e-05,
+      "loss": 3.9092,
+      "step": 1573376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4231047922414262e-05,
+      "loss": 3.892,
+      "step": 1573888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.422267835370747e-05,
+      "loss": 3.9078,
+      "step": 1574400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4214292406196955e-05,
+      "loss": 3.9188,
+      "step": 1574912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4205906458686435e-05,
+      "loss": 3.9015,
+      "step": 1575424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4197520511175915e-05,
+      "loss": 3.8884,
+      "step": 1575936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4189134563665395e-05,
+      "loss": 3.9043,
+      "step": 1576448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4180748616154875e-05,
+      "loss": 3.9061,
+      "step": 1576960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4172362668644355e-05,
+      "loss": 3.9118,
+      "step": 1577472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.416397672113383e-05,
+      "loss": 3.9056,
+      "step": 1577984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.415559077362331e-05,
+      "loss": 3.8997,
+      "step": 1578496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4147221204916527e-05,
+      "loss": 3.8998,
+      "step": 1579008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4138851636209736e-05,
+      "loss": 3.9089,
+      "step": 1579520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4130465688699216e-05,
+      "loss": 3.8896,
+      "step": 1580032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4122079741188696e-05,
+      "loss": 3.9014,
+      "step": 1580544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4113693793678176e-05,
+      "loss": 3.9014,
+      "step": 1581056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.410532422497139e-05,
+      "loss": 3.9082,
+      "step": 1581568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.409693827746087e-05,
+      "loss": 3.8848,
+      "step": 1582080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.408855232995035e-05,
+      "loss": 3.9033,
+      "step": 1582592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4080166382439828e-05,
+      "loss": 3.8951,
+      "step": 1583104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4071780434929305e-05,
+      "loss": 3.8986,
+      "step": 1583616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4063394487418785e-05,
+      "loss": 3.9004,
+      "step": 1584128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4055024918712e-05,
+      "loss": 3.8944,
+      "step": 1584640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.404663897120148e-05,
+      "loss": 3.9054,
+      "step": 1585152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4038253023690957e-05,
+      "loss": 3.91,
+      "step": 1585664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.402988345498417e-05,
+      "loss": 3.8971,
+      "step": 1586176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.402149750747365e-05,
+      "loss": 3.9064,
+      "step": 1586688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4013111559963126e-05,
+      "loss": 3.8976,
+      "step": 1587200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.400472561245261e-05,
+      "loss": 3.9086,
+      "step": 1587712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.399633966494209e-05,
+      "loss": 3.8892,
+      "step": 1588224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.398795371743157e-05,
+      "loss": 3.9114,
+      "step": 1588736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.397956776992105e-05,
+      "loss": 3.8919,
+      "step": 1589248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.397118182241053e-05,
+      "loss": 3.9051,
+      "step": 1589760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.396279587490001e-05,
+      "loss": 3.895,
+      "step": 1590272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.395442630619322e-05,
+      "loss": 3.913,
+      "step": 1590784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3946040358682702e-05,
+      "loss": 3.8996,
+      "step": 1591296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3937654411172182e-05,
+      "loss": 3.8986,
+      "step": 1591808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.392928484246539e-05,
+      "loss": 3.8886,
+      "step": 1592320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.392089889495487e-05,
+      "loss": 3.9161,
+      "step": 1592832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.391251294744435e-05,
+      "loss": 3.903,
+      "step": 1593344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.390412699993383e-05,
+      "loss": 3.905,
+      "step": 1593856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.389574105242331e-05,
+      "loss": 3.8938,
+      "step": 1594368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.388735510491279e-05,
+      "loss": 3.9092,
+      "step": 1594880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.387896915740227e-05,
+      "loss": 3.9125,
+      "step": 1595392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.387058320989175e-05,
+      "loss": 3.8973,
+      "step": 1595904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.386219726238123e-05,
+      "loss": 3.9002,
+      "step": 1596416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.385382769367444e-05,
+      "loss": 3.9083,
+      "step": 1596928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3845441746163923e-05,
+      "loss": 3.9094,
+      "step": 1597440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3837055798653403e-05,
+      "loss": 3.9103,
+      "step": 1597952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3828669851142883e-05,
+      "loss": 3.894,
+      "step": 1598464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3820283903632363e-05,
+      "loss": 3.9045,
+      "step": 1598976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3811914334925572e-05,
+      "loss": 3.9005,
+      "step": 1599488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3803528387415052e-05,
+      "loss": 3.902,
+      "step": 1600000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3795142439904532e-05,
+      "loss": 3.909,
+      "step": 1600512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3786756492394012e-05,
+      "loss": 3.8965,
+      "step": 1601024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3778370544883492e-05,
+      "loss": 3.9096,
+      "step": 1601536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.376998459737297e-05,
+      "loss": 3.8987,
+      "step": 1602048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.376159864986245e-05,
+      "loss": 3.9089,
+      "step": 1602560
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.046624183654785,
+      "eval_runtime": 288.4045,
+      "eval_samples_per_second": 1323.11,
+      "eval_steps_per_second": 41.348,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375321270235193e-05,
+      "loss": 3.8929,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.374482675484141e-05,
+      "loss": 3.9092,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.373644080733089e-05,
+      "loss": 3.909,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.372805485982037e-05,
+      "loss": 3.9041,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371966891230985e-05,
+      "loss": 3.9121,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371128296479933e-05,
+      "loss": 3.897,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370289701728881e-05,
+      "loss": 3.8996,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369451106977829e-05,
+      "loss": 3.8924,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.368612512226777e-05,
+      "loss": 3.9097,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.367773917475725e-05,
+      "loss": 3.9085,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3669353227246728e-05,
+      "loss": 3.9048,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.366096727973621e-05,
+      "loss": 3.9071,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3652597711029424e-05,
+      "loss": 3.8984,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3644211763518904e-05,
+      "loss": 3.8949,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.363582581600838e-05,
+      "loss": 3.8963,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.362743986849786e-05,
+      "loss": 3.8872,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.361905392098734e-05,
+      "loss": 3.8982,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.361066797347682e-05,
+      "loss": 3.8928,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3602282025966303e-05,
+      "loss": 3.8935,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3593896078455783e-05,
+      "loss": 3.9293,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3585510130945263e-05,
+      "loss": 3.895,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357712418343474e-05,
+      "loss": 3.9071,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356873823592422e-05,
+      "loss": 3.905,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.35603522884137e-05,
+      "loss": 3.9036,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3551999098510645e-05,
+      "loss": 3.8963,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3543613151000125e-05,
+      "loss": 3.8966,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3535227203489605e-05,
+      "loss": 3.9031,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3526841255979085e-05,
+      "loss": 3.8872,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351845530846856e-05,
+      "loss": 3.8941,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351006936095804e-05,
+      "loss": 3.893,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3501699792251257e-05,
+      "loss": 3.8973,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3493313844740737e-05,
+      "loss": 3.904,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3484927897230213e-05,
+      "loss": 3.9029,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3476541949719693e-05,
+      "loss": 3.8977,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3468156002209173e-05,
+      "loss": 3.9019,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459770054698653e-05,
+      "loss": 3.906,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3451400485991866e-05,
+      "loss": 3.8928,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3443014538481346e-05,
+      "loss": 3.9071,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434628590970826e-05,
+      "loss": 3.8856,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3426242643460306e-05,
+      "loss": 3.9061,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3417856695949786e-05,
+      "loss": 3.8851,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3409470748439266e-05,
+      "loss": 3.9197,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3401084800928742e-05,
+      "loss": 3.904,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3392698853418226e-05,
+      "loss": 3.8988,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3384329284711438e-05,
+      "loss": 3.9025,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3375959716004647e-05,
+      "loss": 3.8999,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367573768494127e-05,
+      "loss": 3.9018,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3359187820983607e-05,
+      "loss": 3.8907,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3350801873473087e-05,
+      "loss": 3.8824,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3342415925962567e-05,
+      "loss": 3.9128,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3334029978452047e-05,
+      "loss": 3.8987,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325644030941527e-05,
+      "loss": 3.906,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3317258083431007e-05,
+      "loss": 3.8911,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3308888514724216e-05,
+      "loss": 3.8975,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3300502567213696e-05,
+      "loss": 3.891,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3292132998506912e-05,
+      "loss": 3.9005,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328374705099639e-05,
+      "loss": 3.8954,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3275361103485868e-05,
+      "loss": 3.9112,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3266975155975348e-05,
+      "loss": 3.9071,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258589208464828e-05,
+      "loss": 3.8901,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3250203260954308e-05,
+      "loss": 3.8959,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241817313443788e-05,
+      "loss": 3.9051,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.323343136593327e-05,
+      "loss": 3.8868,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.322506179722648e-05,
+      "loss": 3.895,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321669222851969e-05,
+      "loss": 3.8986,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.320830628100917e-05,
+      "loss": 3.9027,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.319992033349865e-05,
+      "loss": 3.8833,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3191534385988133e-05,
+      "loss": 3.8898,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3183148438477613e-05,
+      "loss": 3.8852,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3174762490967093e-05,
+      "loss": 3.9049,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316637654345657e-05,
+      "loss": 3.9001,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.315799059594605e-05,
+      "loss": 3.8961,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314960464843553e-05,
+      "loss": 3.903,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3141251458532474e-05,
+      "loss": 3.8931,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132865511021954e-05,
+      "loss": 3.9135,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3124479563511434e-05,
+      "loss": 3.8859,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3116093616000914e-05,
+      "loss": 3.9053,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3107707668490394e-05,
+      "loss": 3.9004,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309932172097987e-05,
+      "loss": 3.899,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309093577346935e-05,
+      "loss": 3.9023,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3082566204762566e-05,
+      "loss": 3.8971,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3074180257252043e-05,
+      "loss": 3.9027,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3065794309741523e-05,
+      "loss": 3.9048,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3057408362231003e-05,
+      "loss": 3.8866,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3049022414720483e-05,
+      "loss": 3.9043,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3040636467209963e-05,
+      "loss": 3.8933,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3032250519699443e-05,
+      "loss": 3.9044,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3023864572188926e-05,
+      "loss": 3.8952,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3015495003482135e-05,
+      "loss": 3.8888,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3007109055971615e-05,
+      "loss": 3.8976,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2998723108461095e-05,
+      "loss": 3.8969,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2990337160950575e-05,
+      "loss": 3.8989,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2981951213440052e-05,
+      "loss": 3.8834,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2973581644733268e-05,
+      "loss": 3.9041,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2965212076026477e-05,
+      "loss": 3.909,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956826128515957e-05,
+      "loss": 3.8965,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2948440181005437e-05,
+      "loss": 3.8795,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2940054233494917e-05,
+      "loss": 3.8955,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2931668285984397e-05,
+      "loss": 3.8967,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2923282338473876e-05,
+      "loss": 3.9009,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2914896390963356e-05,
+      "loss": 3.9028,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2906510443452836e-05,
+      "loss": 3.893,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.289814087474605e-05,
+      "loss": 3.8917,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889771306039258e-05,
+      "loss": 3.9011,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.288138535852874e-05,
+      "loss": 3.8817,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.287301578982195e-05,
+      "loss": 3.8952,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.286462984231143e-05,
+      "loss": 3.889,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.285624389480091e-05,
+      "loss": 3.9008,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.284785794729039e-05,
+      "loss": 3.8789,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.283947199977987e-05,
+      "loss": 3.8919,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.283108605226935e-05,
+      "loss": 3.8923,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.282270010475883e-05,
+      "loss": 3.89,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.281431415724831e-05,
+      "loss": 3.8933,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.280592820973779e-05,
+      "loss": 3.8833,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.279754226222727e-05,
+      "loss": 3.8992,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.278915631471675e-05,
+      "loss": 3.9005,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2780803124813695e-05,
+      "loss": 3.889,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.277241717730317e-05,
+      "loss": 3.9012,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.276403122979265e-05,
+      "loss": 3.8886,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.275564528228213e-05,
+      "loss": 3.9032,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274725933477161e-05,
+      "loss": 3.8863,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273887338726109e-05,
+      "loss": 3.8987,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273048743975057e-05,
+      "loss": 3.8829,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2722117871043784e-05,
+      "loss": 3.8956,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713731923533264e-05,
+      "loss": 3.8918,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2705345976022744e-05,
+      "loss": 3.9005,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2696960028512224e-05,
+      "loss": 3.8901,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2688574081001704e-05,
+      "loss": 3.8941,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268018813349118e-05,
+      "loss": 3.8771,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.267180218598066e-05,
+      "loss": 3.9098,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2663416238470143e-05,
+      "loss": 3.8916,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2655030290959623e-05,
+      "loss": 3.8984,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2646644343449103e-05,
+      "loss": 3.8849,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2638258395938583e-05,
+      "loss": 3.9031,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262987244842806e-05,
+      "loss": 3.905,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262148650091754e-05,
+      "loss": 3.8912,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2613133311014485e-05,
+      "loss": 3.8948,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604747363503965e-05,
+      "loss": 3.8991,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2596361415993445e-05,
+      "loss": 3.8997,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587975468482925e-05,
+      "loss": 3.9087,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2579589520972405e-05,
+      "loss": 3.8814,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257120357346188e-05,
+      "loss": 3.9016,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2562817625951365e-05,
+      "loss": 3.8893,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2554431678440845e-05,
+      "loss": 3.8985,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2546062109734057e-05,
+      "loss": 3.8983,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537676162223534e-05,
+      "loss": 3.8848,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2529290214713014e-05,
+      "loss": 3.9053,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2520904267202494e-05,
+      "loss": 3.8953,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2512534698495706e-05,
+      "loss": 3.8981,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.043795108795166,
+      "eval_runtime": 308.5427,
+      "eval_samples_per_second": 1236.753,
+      "eval_steps_per_second": 38.649,
+      "step": 1679040
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.466866265568689e+17,
+  "trial_name": null,
+  "trial_params": null
+}