diff --git "a/checkpoint-1373760/trainer_state.json" "b/checkpoint-1373760/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1373760/trainer_state.json"
@@ -0,0 +1,16267 @@
+{
+  "best_metric": 4.00921106338501,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/lstm/3/checkpoints/checkpoint-1373760",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1373760,
+  "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": 10.8199,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5528,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.06,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9942,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9475,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9053,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7338,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6207,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5238,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4509,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.388,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3338,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2787,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2174,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1579,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0959,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0476,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0002,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9599,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9202,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8865,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8514,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.821,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7926,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7751,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.736,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7149,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6914,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6697,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6433,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6201,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6143,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.587,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5777,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5633,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.539,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5288,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4888,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4896,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4544,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4537,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4357,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 5.4352,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 5.4078,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.401,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9622763392456454e-05,
+      "loss": 5.3937,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9614377444945934e-05,
+      "loss": 5.3759,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 5.3796,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 5.3508,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3354,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3452,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3374,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3114,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.3008,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2992,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.283,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2967,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2575,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2568,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2493,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2378,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2388,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2172,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2114,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2033,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2168,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1882,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1869,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1608,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.168,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1779,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1684,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1447,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1598,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.155,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1314,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1282,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1113,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0988,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0973,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1008,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0977,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.1055,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0902,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.07,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0708,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.072,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0601,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0531,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.041,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0406,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0424,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.03,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0317,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0198,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0152,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0036,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9995,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.994,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9961,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9985,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9754,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.981,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9702,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9669,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.959,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9539,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9525,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9442,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9417,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9341,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9246,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9353,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.924,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.925,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9197,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9203,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9114,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.8966,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.905,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9096,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8952,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8879,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.886,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8856,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.874,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8772,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8674,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8691,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8572,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8777,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8588,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.852,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8466,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8477,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8483,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8492,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8371,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8388,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8336,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8317,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8189,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8307,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8176,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8042,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8188,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8067,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8074,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8137,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.793,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.761042594909668,
+      "eval_runtime": 291.6957,
+      "eval_samples_per_second": 1308.182,
+      "eval_steps_per_second": 40.882,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7875,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.7901,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.7919,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.7869,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7908,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7758,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7823,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7674,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7763,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7651,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7626,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7653,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7577,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.7524,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.7493,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.7363,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.7464,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7386,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7326,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7513,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7331,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7356,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7295,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7468,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7174,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7181,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7228,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7222,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7083,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.7083,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.7173,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.7149,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.717,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7103,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.7013,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7084,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.681,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6998,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.6757,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.689,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6807,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6928,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6812,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6778,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6804,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.6727,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6859,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6621,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6556,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6763,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.6753,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315030566123526e-05,
+      "loss": 4.6598,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.6526,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.6598,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.647,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.6699,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6443,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6527,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.646,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6383,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6499,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6347,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.6314,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6324,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6474,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6299,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.627,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.62,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6217,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6413,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6373,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814737713112805e-05,
+      "loss": 4.6177,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6319,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6364,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6223,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.6241,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.6117,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.5997,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808867549855441e-05,
+      "loss": 4.6036,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6075,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.6111,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6286,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.6138,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.5956,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.603,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.6056,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6031,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.594,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.5931,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5915,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5981,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5875,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5942,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5886,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5871,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5785,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.584,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5761,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5813,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5853,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5705,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5749,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5692,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.5712,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5628,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5639,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.567,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5578,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5616,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828809398550675e-05,
+      "loss": 4.554,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820423451040155e-05,
+      "loss": 4.5476,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812037503529635e-05,
+      "loss": 4.5687,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.5539,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5592,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.554,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.564,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770124144780764e-05,
+      "loss": 4.5554,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761738197270244e-05,
+      "loss": 4.541,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.5523,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774498268105293e-05,
+      "loss": 4.5578,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773659673354241e-05,
+      "loss": 4.5424,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5468,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5402,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5479,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5358,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5419,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.535,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5392,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766954191106572e-05,
+      "loss": 4.5262,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76611559635552e-05,
+      "loss": 4.5529,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765277001604468e-05,
+      "loss": 4.5403,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764438406853416e-05,
+      "loss": 4.5257,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.528,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762761217351312e-05,
+      "loss": 4.5311,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76192262260026e-05,
+      "loss": 4.5341,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761084027849208e-05,
+      "loss": 4.5422,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760245433098156e-05,
+      "loss": 4.5231,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759406838347104e-05,
+      "loss": 4.5283,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758568243596052e-05,
+      "loss": 4.5248,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757731286725373e-05,
+      "loss": 4.5298,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568926919743206e-05,
+      "loss": 4.5202,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560540972232686e-05,
+      "loss": 4.5283,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552155024722166e-05,
+      "loss": 4.5181,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543785456015375e-05,
+      "loss": 4.5142,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.5256,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.5179,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.516,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5198,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750187209726651e-05,
+      "loss": 4.5078,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.4815473556518555,
+      "eval_runtime": 291.2912,
+      "eval_samples_per_second": 1309.999,
+      "eval_steps_per_second": 40.938,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749348614975599e-05,
+      "loss": 4.5017,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748510020224547e-05,
+      "loss": 4.5019,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.5148,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746832830722443e-05,
+      "loss": 4.5047,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745994235971391e-05,
+      "loss": 4.5173,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.5024,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.5052,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.4973,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.5088,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.4968,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.4976,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.4978,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.493,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.4891,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4909,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4792,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4871,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.4797,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4769,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5043,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4846,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4792,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4829,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4992,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4716,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4765,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4762,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4823,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4661,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4717,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4749,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4803,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.4842,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4749,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4662,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4749,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719167393339592e-05,
+      "loss": 4.4556,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.4695,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4493,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4586,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4594,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4697,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4596,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4555,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4585,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711623316340871e-05,
+      "loss": 4.4547,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710784721589819e-05,
+      "loss": 4.4667,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099461268387667e-05,
+      "loss": 4.448,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091075320877147e-05,
+      "loss": 4.4368,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082689373366627e-05,
+      "loss": 4.4627,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.4573,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4462,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.4443,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049145583324546e-05,
+      "loss": 4.4482,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040759635814026e-05,
+      "loss": 4.4372,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032373688303506e-05,
+      "loss": 4.4576,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7023987740792986e-05,
+      "loss": 4.438,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4439,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4373,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4372,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4398,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4368,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697372119214106e-05,
+      "loss": 4.4288,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.431,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.446,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.434,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4292,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.423,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4268,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4464,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.44,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4269,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4385,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688149214832907e-05,
+      "loss": 4.4432,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687310620081855e-05,
+      "loss": 4.4255,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4318,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.4244,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4136,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.4208,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.4161,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822823273362894e-05,
+      "loss": 4.4246,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.437,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4323,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4087,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.4239,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4192,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.4238,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676413801959298e-05,
+      "loss": 4.4147,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675575207208246e-05,
+      "loss": 4.4101,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.4152,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4129,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673062698715837e-05,
+      "loss": 4.4107,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672224103964785e-05,
+      "loss": 4.4166,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.4105,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4156,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.4057,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668869724960576e-05,
+      "loss": 4.4047,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.4042,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667192535458472e-05,
+      "loss": 4.4123,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.412,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655169838367416e-05,
+      "loss": 4.3975,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4019,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.3994,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.4031,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.3916,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661327285842228e-05,
+      "loss": 4.3922,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660488691091176e-05,
+      "loss": 4.4037,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596500963401234e-05,
+      "loss": 4.3951,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588115015890714e-05,
+      "loss": 4.3927,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579729068380194e-05,
+      "loss": 4.3898,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571343120869674e-05,
+      "loss": 4.3845,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.3998,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655458760465237e-05,
+      "loss": 4.3944,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.392,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.3934,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.404,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.395,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.3815,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.392,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649588597207873e-05,
+      "loss": 4.3955,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648751640337194e-05,
+      "loss": 4.3858,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647914683466515e-05,
+      "loss": 4.3894,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.3823,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.3931,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.3775,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64456194234268e-05,
+      "loss": 4.3918,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643723347591628e-05,
+      "loss": 4.3722,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642886390720949e-05,
+      "loss": 4.389,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642047795969897e-05,
+      "loss": 4.372,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641209201218845e-05,
+      "loss": 4.3981,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640370606467793e-05,
+      "loss": 4.3895,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639532011716741e-05,
+      "loss": 4.3794,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638693416965689e-05,
+      "loss": 4.3715,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637854822214637e-05,
+      "loss": 4.3819,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637016227463585e-05,
+      "loss": 4.3813,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636177632712533e-05,
+      "loss": 4.3922,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.3726,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634500443210429e-05,
+      "loss": 4.3827,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.3775,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.3865,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.3704,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.3804,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.3737,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62947051258449e-05,
+      "loss": 4.3671,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628633555713812e-05,
+      "loss": 4.3798,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277949609627597e-05,
+      "loss": 4.3774,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626956366211707e-05,
+      "loss": 4.3669,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261194093410286e-05,
+      "loss": 4.3774,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252808145899766e-05,
+      "loss": 4.3662,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.345434665679932,
+      "eval_runtime": 295.7775,
+      "eval_samples_per_second": 1290.129,
+      "eval_steps_per_second": 40.317,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244422198389246e-05,
+      "loss": 4.3643,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236036250878726e-05,
+      "loss": 4.3587,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227650303368205e-05,
+      "loss": 4.3707,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3678,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.38,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3598,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.366,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3579,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177334618305085e-05,
+      "loss": 4.376,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3552,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3576,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3646,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3549,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3541,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.3543,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3476,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3477,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.3484,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3436,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3716,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3495,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3491,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3514,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605159454205848e-05,
+      "loss": 4.3638,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3439,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.346,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602646945713438e-05,
+      "loss": 4.3453,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018083509623864e-05,
+      "loss": 4.3502,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009697562113344e-05,
+      "loss": 4.3387,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001311614602824e-05,
+      "loss": 4.3432,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5992925667092303e-05,
+      "loss": 4.3461,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984539719581783e-05,
+      "loss": 4.3514,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976153772071257e-05,
+      "loss": 4.3566,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967767824560737e-05,
+      "loss": 4.3475,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959381877050217e-05,
+      "loss": 4.3418,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5950995929539696e-05,
+      "loss": 4.3457,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942609982029176e-05,
+      "loss": 4.3306,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934240413322386e-05,
+      "loss": 4.3404,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5925854465811865e-05,
+      "loss": 4.3271,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917468518301345e-05,
+      "loss": 4.3318,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590908257079083e-05,
+      "loss": 4.3346,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590069662328031e-05,
+      "loss": 4.3484,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589231067576979e-05,
+      "loss": 4.341,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588392472825927e-05,
+      "loss": 4.3306,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587553878074875e-05,
+      "loss": 4.3353,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3319,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585878326453144e-05,
+      "loss": 4.3433,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585039731702092e-05,
+      "loss": 4.3267,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58420113695104e-05,
+      "loss": 4.3164,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3399,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582525585329309e-05,
+      "loss": 4.3366,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581686990578257e-05,
+      "loss": 4.3217,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580848395827205e-05,
+      "loss": 4.329,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800114389565266e-05,
+      "loss": 4.3314,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3155,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3356,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.317,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3298,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3172,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3274,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741429135795355e-05,
+      "loss": 4.3181,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3228,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724673619578044e-05,
+      "loss": 4.3102,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.3178,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5707901724557004e-05,
+      "loss": 4.3339,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699515777046484e-05,
+      "loss": 4.3156,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.316,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682743882025443e-05,
+      "loss": 4.3074,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.3152,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3335,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565760241829762e-05,
+      "loss": 4.3258,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.3193,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564083052327657e-05,
+      "loss": 4.3266,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563244457576605e-05,
+      "loss": 4.3277,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562405862825553e-05,
+      "loss": 4.3158,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561567268074501e-05,
+      "loss": 4.3191,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3166,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3059,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.3078,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3062,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557375932199615e-05,
+      "loss": 4.3136,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3261,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3231,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.2976,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55402319107578e-05,
+      "loss": 4.3212,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553184596324728e-05,
+      "loss": 4.3113,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3132,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3076,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.3014,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.3101,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.3055,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.3057,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54731770882811e-05,
+      "loss": 4.3086,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.3039,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.3125,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3001,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543963329823902e-05,
+      "loss": 4.2996,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.2982,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.3108,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.3055,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540610588700067e-05,
+      "loss": 4.2949,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539771993949015e-05,
+      "loss": 4.2964,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538935037078336e-05,
+      "loss": 4.2999,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538096442327284e-05,
+      "loss": 4.3039,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372578475762326e-05,
+      "loss": 4.2931,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364192528251806e-05,
+      "loss": 4.2878,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.3021,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.2899,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533903468572024e-05,
+      "loss": 4.2909,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.287,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532227916950293e-05,
+      "loss": 4.2889,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5313909600796144e-05,
+      "loss": 4.3005,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2874,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2917,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.2953,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280365810754064e-05,
+      "loss": 4.3029,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527199624204728e-05,
+      "loss": 4.3019,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526361029453675e-05,
+      "loss": 4.2818,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525522434702623e-05,
+      "loss": 4.2927,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524683839951571e-05,
+      "loss": 4.2932,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523846883080893e-05,
+      "loss": 4.2909,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52300828832984e-05,
+      "loss": 4.2937,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522169693578788e-05,
+      "loss": 4.2833,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521331098827736e-05,
+      "loss": 4.2952,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520492504076684e-05,
+      "loss": 4.2823,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519653909325632e-05,
+      "loss": 4.2916,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2791,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517978357703902e-05,
+      "loss": 4.294,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517141400833223e-05,
+      "loss": 4.2739,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516302806082171e-05,
+      "loss": 4.306,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154642113311187e-05,
+      "loss": 4.2863,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146256165800667e-05,
+      "loss": 4.2871,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137870218290147e-05,
+      "loss": 4.2768,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129500649583356e-05,
+      "loss": 4.2896,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121114702072836e-05,
+      "loss": 4.2877,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112728754562316e-05,
+      "loss": 4.297,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104342807051795e-05,
+      "loss": 4.2778,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095956859541275e-05,
+      "loss": 4.2877,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087570912030755e-05,
+      "loss": 4.2843,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507920134332397e-05,
+      "loss": 4.2949,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507081539581345e-05,
+      "loss": 4.274,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506242944830293e-05,
+      "loss": 4.2876,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505404350079241e-05,
+      "loss": 4.2802,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504565755328189e-05,
+      "loss": 4.2793,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503727160577137e-05,
+      "loss": 4.2863,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2844,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502051608955406e-05,
+      "loss": 4.2747,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501213014204354e-05,
+      "loss": 4.285,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500374419453302e-05,
+      "loss": 4.2772,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.262485504150391,
+      "eval_runtime": 312.9738,
+      "eval_samples_per_second": 1219.243,
+      "eval_steps_per_second": 38.102,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49953582470225e-05,
+      "loss": 4.272,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498697229951198e-05,
+      "loss": 4.266,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497858635200146e-05,
+      "loss": 4.2805,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2768,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.29,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2733,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2746,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493665661444886e-05,
+      "loss": 4.2707,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492827066693834e-05,
+      "loss": 4.2845,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491988471942782e-05,
+      "loss": 4.2714,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49114987719173e-05,
+      "loss": 4.2678,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490311282440678e-05,
+      "loss": 4.2726,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489474325569999e-05,
+      "loss": 4.2666,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488635730818947e-05,
+      "loss": 4.2674,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487797136067895e-05,
+      "loss": 4.2715,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486958541316843e-05,
+      "loss": 4.2545,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486119946565791e-05,
+      "loss": 4.2624,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485281351814739e-05,
+      "loss": 4.2618,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484442757063687e-05,
+      "loss": 4.2541,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483604162312635e-05,
+      "loss": 4.2806,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482765567561583e-05,
+      "loss": 4.2673,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481926972810531e-05,
+      "loss": 4.263,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481088378059479e-05,
+      "loss": 4.2649,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480249783308427e-05,
+      "loss": 4.2777,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479412826437748e-05,
+      "loss": 4.2568,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478574231686696e-05,
+      "loss": 4.2627,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477735636935644e-05,
+      "loss": 4.2595,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476898680064965e-05,
+      "loss": 4.2615,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760600853139126e-05,
+      "loss": 4.2545,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752214905628606e-05,
+      "loss": 4.2602,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743828958118086e-05,
+      "loss": 4.2625,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473544301060757e-05,
+      "loss": 4.2667,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472705706309705e-05,
+      "loss": 4.2724,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471867111558653e-05,
+      "loss": 4.2653,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471028516807601e-05,
+      "loss": 4.2561,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470193197817295e-05,
+      "loss": 4.2623,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469354603066243e-05,
+      "loss": 4.2553,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468516008315191e-05,
+      "loss": 4.2567,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467677413564139e-05,
+      "loss": 4.2446,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466838818813087e-05,
+      "loss": 4.2506,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466001861942408e-05,
+      "loss": 4.2545,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465163267191356e-05,
+      "loss": 4.2656,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464324672440304e-05,
+      "loss": 4.2579,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634860776892526e-05,
+      "loss": 4.252,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626474829382006e-05,
+      "loss": 4.2534,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618088881871486e-05,
+      "loss": 4.2509,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609702934360966e-05,
+      "loss": 4.261,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601316986850446e-05,
+      "loss": 4.2478,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4592947418143655e-05,
+      "loss": 4.2344,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584561470633135e-05,
+      "loss": 4.259,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576175523122615e-05,
+      "loss": 4.2573,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567789575612095e-05,
+      "loss": 4.24,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2475,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551034059394784e-05,
+      "loss": 4.2538,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542648111884264e-05,
+      "loss": 4.2349,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534262164373744e-05,
+      "loss": 4.2571,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525876216863224e-05,
+      "loss": 4.2402,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517490269352704e-05,
+      "loss": 4.2506,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450910432184219e-05,
+      "loss": 4.2388,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45007347531354e-05,
+      "loss": 4.2455,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449234880562488e-05,
+      "loss": 4.2416,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448396285811436e-05,
+      "loss": 4.2475,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447559328940757e-05,
+      "loss": 4.2304,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446722372070078e-05,
+      "loss": 4.2432,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445883777319026e-05,
+      "loss": 4.2536,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445045182567974e-05,
+      "loss": 4.2416,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444206587816922e-05,
+      "loss": 4.2371,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44336799306587e-05,
+      "loss": 4.2298,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442529398314818e-05,
+      "loss": 4.2361,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441690803563766e-05,
+      "loss": 4.2531,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440852208812714e-05,
+      "loss": 4.25,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440013614061662e-05,
+      "loss": 4.2443,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43917501931061e-05,
+      "loss": 4.2512,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438336424559558e-05,
+      "loss": 4.255,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437499467688879e-05,
+      "loss": 4.2385,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436660872937827e-05,
+      "loss": 4.2418,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358222781867747e-05,
+      "loss": 4.2436,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349836834357227e-05,
+      "loss": 4.2317,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341450886846707e-05,
+      "loss": 4.2318,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333081318139916e-05,
+      "loss": 4.2346,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324695370629396e-05,
+      "loss": 4.2363,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431630942311888e-05,
+      "loss": 4.2524,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430792347560836e-05,
+      "loss": 4.2504,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429955390690157e-05,
+      "loss": 4.225,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429116795939105e-05,
+      "loss": 4.242,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428278201188053e-05,
+      "loss": 4.2389,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427439606437001e-05,
+      "loss": 4.2389,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426602649566322e-05,
+      "loss": 4.2364,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42576405481527e-05,
+      "loss": 4.2268,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424925460064218e-05,
+      "loss": 4.2365,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424086865313166e-05,
+      "loss": 4.2336,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423248270562114e-05,
+      "loss": 4.2321,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422409675811062e-05,
+      "loss": 4.2315,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42157108106001e-05,
+      "loss": 4.2348,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420732486308958e-05,
+      "loss": 4.2355,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419893891557907e-05,
+      "loss": 4.2342,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190569346872276e-05,
+      "loss": 4.2228,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182183399361756e-05,
+      "loss": 4.2293,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4173797451851236e-05,
+      "loss": 4.234,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165411504340716e-05,
+      "loss": 4.2379,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157041935633925e-05,
+      "loss": 4.222,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148655988123405e-05,
+      "loss": 4.2266,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140270040612885e-05,
+      "loss": 4.2291,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4131884093102365e-05,
+      "loss": 4.2317,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123514524395574e-05,
+      "loss": 4.2236,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115128576885054e-05,
+      "loss": 4.2193,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106742629374534e-05,
+      "loss": 4.2261,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409835668186402e-05,
+      "loss": 4.223,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899707343535e-05,
+      "loss": 4.2183,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081584786842974e-05,
+      "loss": 4.2193,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073198839332453e-05,
+      "loss": 4.2179,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064812891821933e-05,
+      "loss": 4.231,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056426944311413e-05,
+      "loss": 4.2217,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404805737560462e-05,
+      "loss": 4.2197,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40396714280941e-05,
+      "loss": 4.2258,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403128548058358e-05,
+      "loss": 4.2363,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402289953307306e-05,
+      "loss": 4.2342,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401452996436627e-05,
+      "loss": 4.215,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400614401685576e-05,
+      "loss": 4.2214,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399777444814897e-05,
+      "loss": 4.225,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398938850063845e-05,
+      "loss": 4.2199,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398100255312793e-05,
+      "loss": 4.2253,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397261660561741e-05,
+      "loss": 4.2144,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396423065810689e-05,
+      "loss": 4.2241,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395584471059637e-05,
+      "loss": 4.2209,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394745876308585e-05,
+      "loss": 4.2228,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393907281557533e-05,
+      "loss": 4.2097,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930703246868536e-05,
+      "loss": 4.2215,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922317299358016e-05,
+      "loss": 4.2157,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913931351847496e-05,
+      "loss": 4.2301,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905545404336976e-05,
+      "loss": 4.2197,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897159456826456e-05,
+      "loss": 4.224,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888773509315936e-05,
+      "loss": 4.2086,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388038756180542e-05,
+      "loss": 4.2245,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38720016142949e-05,
+      "loss": 4.2223,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386363204558811e-05,
+      "loss": 4.2264,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385524609807759e-05,
+      "loss": 4.2107,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384686015056707e-05,
+      "loss": 4.2275,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383849058186028e-05,
+      "loss": 4.2141,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383010463434976e-05,
+      "loss": 4.2308,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382171868683924e-05,
+      "loss": 4.21,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381333273932872e-05,
+      "loss": 4.2216,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38049467918182e-05,
+      "loss": 4.2142,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379656084430768e-05,
+      "loss": 4.2095,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378819127560089e-05,
+      "loss": 4.2219,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779805328090376e-05,
+      "loss": 4.2181,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771419380579856e-05,
+      "loss": 4.2108,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763033433069336e-05,
+      "loss": 4.2166,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375464748555881e-05,
+      "loss": 4.2165,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.206160068511963,
+      "eval_runtime": 299.6313,
+      "eval_samples_per_second": 1273.535,
+      "eval_steps_per_second": 39.799,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.2095,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373787559053777e-05,
+      "loss": 4.2013,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372948964302725e-05,
+      "loss": 4.2138,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372110369551673e-05,
+      "loss": 4.2115,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371271774800621e-05,
+      "loss": 4.2227,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370433180049569e-05,
+      "loss": 4.2095,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369594585298517e-05,
+      "loss": 4.2086,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368755990547465e-05,
+      "loss": 4.2079,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367917395796413e-05,
+      "loss": 4.219,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367078801045361e-05,
+      "loss": 4.2069,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366240206294309e-05,
+      "loss": 4.2075,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3654016115432576e-05,
+      "loss": 4.2066,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3645646546725785e-05,
+      "loss": 4.2053,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3637260599215265e-05,
+      "loss": 4.2047,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3628874651704745e-05,
+      "loss": 4.208,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3620488704194225e-05,
+      "loss": 4.1932,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36121027566837e-05,
+      "loss": 4.1979,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360371680917318e-05,
+      "loss": 4.201,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359533086166266e-05,
+      "loss": 4.1926,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358694491415214e-05,
+      "loss": 4.2157,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357857534544535e-05,
+      "loss": 4.2096,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357020577673856e-05,
+      "loss": 4.2031,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.1985,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355345026052126e-05,
+      "loss": 4.2126,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354506431301074e-05,
+      "loss": 4.1981,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353667836550022e-05,
+      "loss": 4.2034,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352829241798969e-05,
+      "loss": 4.1952,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351990647047917e-05,
+      "loss": 4.2016,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351152052296865e-05,
+      "loss": 4.1972,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350313457545813e-05,
+      "loss": 4.1978,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349474862794761e-05,
+      "loss": 4.2007,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348636268043709e-05,
+      "loss": 4.2055,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347797673292657e-05,
+      "loss": 4.2091,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.2056,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.1948,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345281889039501e-05,
+      "loss": 4.2034,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344444932168823e-05,
+      "loss": 4.1963,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343606337417771e-05,
+      "loss": 4.1873,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342767742666719e-05,
+      "loss": 4.1903,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341929147915667e-05,
+      "loss": 4.1868,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341090553164615e-05,
+      "loss": 4.1941,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402535962939356e-05,
+      "loss": 4.2087,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394150015428836e-05,
+      "loss": 4.1942,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385764067918316e-05,
+      "loss": 4.195,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.1996,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3368992172897276e-05,
+      "loss": 4.1916,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360622604190485e-05,
+      "loss": 4.198,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352236656679965e-05,
+      "loss": 4.1909,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334386708797318e-05,
+      "loss": 4.1752,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333548114046266e-05,
+      "loss": 4.202,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332709519295214e-05,
+      "loss": 4.1942,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331870924544162e-05,
+      "loss": 4.1878,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1866,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.1909,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.1775,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285165455399534e-05,
+      "loss": 4.1944,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327679588669275e-05,
+      "loss": 4.1866,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.1928,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.1831,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.1857,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324325209665067e-05,
+      "loss": 4.1826,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323486614914015e-05,
+      "loss": 4.193,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322648020162963e-05,
+      "loss": 4.176,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321809425411911e-05,
+      "loss": 4.1845,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320970830660859e-05,
+      "loss": 4.1944,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320132235909807e-05,
+      "loss": 4.1867,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319293641158755e-05,
+      "loss": 4.1797,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318456684288076e-05,
+      "loss": 4.1782,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317618089537024e-05,
+      "loss": 4.1783,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316779494785972e-05,
+      "loss": 4.1931,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31594090003492e-05,
+      "loss": 4.1942,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315102305283868e-05,
+      "loss": 4.1847,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314263710532816e-05,
+      "loss": 4.1934,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313425115781764e-05,
+      "loss": 4.1977,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312586521030712e-05,
+      "loss": 4.1888,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311751202040406e-05,
+      "loss": 4.1819,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.1833,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310074012538302e-05,
+      "loss": 4.182,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30923541778725e-05,
+      "loss": 4.1741,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308398460916571e-05,
+      "loss": 4.1834,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307559866165519e-05,
+      "loss": 4.1766,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306721271414467e-05,
+      "loss": 4.1946,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305882676663415e-05,
+      "loss": 4.1972,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305044081912363e-05,
+      "loss": 4.1666,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.1907,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.1779,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30252993553958e-05,
+      "loss": 4.1844,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301691340788529e-05,
+      "loss": 4.1813,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300852746037477e-05,
+      "loss": 4.1733,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000157891667976e-05,
+      "loss": 4.1811,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2991771944157456e-05,
+      "loss": 4.1773,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983385996646936e-05,
+      "loss": 4.179,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975000049136416e-05,
+      "loss": 4.1756,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296661410162589e-05,
+      "loss": 4.1824,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958244532919105e-05,
+      "loss": 4.1791,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949858585408585e-05,
+      "loss": 4.1798,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2941472637898065e-05,
+      "loss": 4.1686,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293308669038754e-05,
+      "loss": 4.177,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2924700742877025e-05,
+      "loss": 4.1766,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2916314795366505e-05,
+      "loss": 4.1855,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2907928847855985e-05,
+      "loss": 4.1696,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2899542900345465e-05,
+      "loss": 4.1759,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2891156952834945e-05,
+      "loss": 4.1702,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2882787384128154e-05,
+      "loss": 4.1799,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287441781542136e-05,
+      "loss": 4.1666,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286603186791084e-05,
+      "loss": 4.1683,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285764592040032e-05,
+      "loss": 4.1704,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28492599728898e-05,
+      "loss": 4.1694,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284089040418301e-05,
+      "loss": 4.1637,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283250445667249e-05,
+      "loss": 4.1707,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282411850916198e-05,
+      "loss": 4.1666,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281573256165146e-05,
+      "loss": 4.1734,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280734661414094e-05,
+      "loss": 4.1729,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279897704543415e-05,
+      "loss": 4.1632,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279059109792363e-05,
+      "loss": 4.1744,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278220515041311e-05,
+      "loss": 4.1855,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277381920290259e-05,
+      "loss": 4.1816,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276543325539207e-05,
+      "loss": 4.1607,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275704730788155e-05,
+      "loss": 4.1693,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274867773917476e-05,
+      "loss": 4.1705,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740291791664237e-05,
+      "loss": 4.1661,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2731905844153717e-05,
+      "loss": 4.1737,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2723519896643196e-05,
+      "loss": 4.1613,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271515032793641e-05,
+      "loss": 4.1715,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270676438042589e-05,
+      "loss": 4.1671,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269837843291537e-05,
+      "loss": 4.1748,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268999248540485e-05,
+      "loss": 4.1609,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268160653789433e-05,
+      "loss": 4.1639,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267322059038381e-05,
+      "loss": 4.169,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266483464287329e-05,
+      "loss": 4.1765,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265644869536277e-05,
+      "loss": 4.1706,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264806274785225e-05,
+      "loss": 4.1704,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639676800341725e-05,
+      "loss": 4.1587,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2631290852831205e-05,
+      "loss": 4.1746,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622904905320685e-05,
+      "loss": 4.1691,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26145353366139e-05,
+      "loss": 4.1764,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260614938910338e-05,
+      "loss": 4.1615,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25977798203966e-05,
+      "loss": 4.1732,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258939387288608e-05,
+      "loss": 4.1645,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258100792537555e-05,
+      "loss": 4.1821,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257262197786503e-05,
+      "loss": 4.158,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564252409158246e-05,
+      "loss": 4.1734,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2555866461647726e-05,
+      "loss": 4.162,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25474805141372e-05,
+      "loss": 4.1598,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253909456662668e-05,
+      "loss": 4.1701,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253070861911616e-05,
+      "loss": 4.1646,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252232267160564e-05,
+      "loss": 4.1675,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251393672409512e-05,
+      "loss": 4.161,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25055507765846e-05,
+      "loss": 4.1704,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.164487361907959,
+      "eval_runtime": 299.0787,
+      "eval_samples_per_second": 1275.888,
+      "eval_steps_per_second": 39.872,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2497181207877815e-05,
+      "loss": 4.1621,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2488811639171024e-05,
+      "loss": 4.1531,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2480425691660504e-05,
+      "loss": 4.1614,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247205612295372e-05,
+      "loss": 4.1632,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24636701754432e-05,
+      "loss": 4.1715,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245528422793267e-05,
+      "loss": 4.165,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244689828042215e-05,
+      "loss": 4.1516,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243851233291163e-05,
+      "loss": 4.1618,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243012638540111e-05,
+      "loss": 4.1666,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242174043789059e-05,
+      "loss": 4.16,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241335449038007e-05,
+      "loss": 4.1583,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240496854286955e-05,
+      "loss": 4.1588,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239659897416277e-05,
+      "loss": 4.1608,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238821302665225e-05,
+      "loss": 4.1538,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237982707914173e-05,
+      "loss": 4.1605,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237144113163121e-05,
+      "loss": 4.1428,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236305518412069e-05,
+      "loss": 4.148,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23546856154139e-05,
+      "loss": 4.1514,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234629966790338e-05,
+      "loss": 4.146,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2337930099196586e-05,
+      "loss": 4.1665,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2329544151686066e-05,
+      "loss": 4.1629,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2321158204175546e-05,
+      "loss": 4.1561,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2312772256665026e-05,
+      "loss": 4.1486,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230440268795824e-05,
+      "loss": 4.1646,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229601674044772e-05,
+      "loss": 4.1504,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22876307929372e-05,
+      "loss": 4.1563,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227924484542668e-05,
+      "loss": 4.1469,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227085889791616e-05,
+      "loss": 4.1545,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226247295040564e-05,
+      "loss": 4.1545,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225408700289512e-05,
+      "loss": 4.1485,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22457010553846e-05,
+      "loss": 4.1528,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223731510787408e-05,
+      "loss": 4.1589,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222892916036356e-05,
+      "loss": 4.1577,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2220543212853035e-05,
+      "loss": 4.1605,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212157265342515e-05,
+      "loss": 4.1472,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220378769663573e-05,
+      "loss": 4.1611,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219540174912521e-05,
+      "loss": 4.1491,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218701580161469e-05,
+      "loss": 4.1399,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217862985410417e-05,
+      "loss": 4.1443,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217024390659365e-05,
+      "loss": 4.1414,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216185795908313e-05,
+      "loss": 4.1482,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215347201157261e-05,
+      "loss": 4.1591,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214508606406209e-05,
+      "loss": 4.1471,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213670011655157e-05,
+      "loss": 4.1492,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212833054784478e-05,
+      "loss": 4.1539,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211994460033426e-05,
+      "loss": 4.1464,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211157503162747e-05,
+      "loss": 4.1522,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210318908411695e-05,
+      "loss": 4.141,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209480313660643e-05,
+      "loss": 4.1301,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208641718909591e-05,
+      "loss": 4.1547,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2078031241585395e-05,
+      "loss": 4.1513,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2069645294074875e-05,
+      "loss": 4.1372,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061259346564355e-05,
+      "loss": 4.1473,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2052889777857564e-05,
+      "loss": 4.1411,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2044503830347044e-05,
+      "loss": 4.1346,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036117882836524e-05,
+      "loss": 4.1455,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027731935326004e-05,
+      "loss": 4.1431,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019345987815484e-05,
+      "loss": 4.1485,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2010960040304964e-05,
+      "loss": 4.1382,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2002574092794444e-05,
+      "loss": 4.1416,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199420452408765e-05,
+      "loss": 4.1378,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198581857657713e-05,
+      "loss": 4.1498,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197743262906661e-05,
+      "loss": 4.1322,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196904668155609e-05,
+      "loss": 4.1387,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196067711284931e-05,
+      "loss": 4.1492,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195229116533879e-05,
+      "loss": 4.1428,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194390521782827e-05,
+      "loss": 4.138,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193551927031775e-05,
+      "loss": 4.1346,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192714970161096e-05,
+      "loss": 4.1303,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191876375410044e-05,
+      "loss": 4.143,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191037780658992e-05,
+      "loss": 4.1535,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19019918590794e-05,
+      "loss": 4.1408,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189360591156887e-05,
+      "loss": 4.1489,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885236342862086e-05,
+      "loss": 4.151,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1876850395351566e-05,
+      "loss": 4.1477,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1868464447841046e-05,
+      "loss": 4.1401,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860078500330526e-05,
+      "loss": 4.1392,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1851692552820006e-05,
+      "loss": 4.1384,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184332298411322e-05,
+      "loss": 4.1274,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1834937036602695e-05,
+      "loss": 4.1416,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1826551089092175e-05,
+      "loss": 4.1366,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818165141581655e-05,
+      "loss": 4.1501,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1809779194071135e-05,
+      "loss": 4.1518,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1801409625364344e-05,
+      "loss": 4.1249,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793023677853824e-05,
+      "loss": 4.1494,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1784637730343304e-05,
+      "loss": 4.1361,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776251782832784e-05,
+      "loss": 4.1418,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767865835322264e-05,
+      "loss": 4.137,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.175949626661548e-05,
+      "loss": 4.1308,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175111031910496e-05,
+      "loss": 4.1396,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174272437159444e-05,
+      "loss": 4.1348,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173433842408392e-05,
+      "loss": 4.1345,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17259524765734e-05,
+      "loss": 4.1347,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171756652906288e-05,
+      "loss": 4.138,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170918058155236e-05,
+      "loss": 4.1433,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170079463404184e-05,
+      "loss": 4.1298,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169242506533505e-05,
+      "loss": 4.1297,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168403911782453e-05,
+      "loss": 4.1321,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167565317031401e-05,
+      "loss": 4.1334,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166726722280349e-05,
+      "loss": 4.1451,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165888127529297e-05,
+      "loss": 4.1299,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165049532778245e-05,
+      "loss": 4.1311,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642109380271935e-05,
+      "loss": 4.1299,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163372343276141e-05,
+      "loss": 4.1398,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162533748525089e-05,
+      "loss": 4.1215,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161698429534783e-05,
+      "loss": 4.1288,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160859834783731e-05,
+      "loss": 4.1288,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160021240032679e-05,
+      "loss": 4.1297,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159182645281627e-05,
+      "loss": 4.1233,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158344050530575e-05,
+      "loss": 4.1293,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1575054557795226e-05,
+      "loss": 4.1232,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156668498908844e-05,
+      "loss": 4.1322,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155829904157792e-05,
+      "loss": 4.1311,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15499130940674e-05,
+      "loss": 4.1248,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154154352536062e-05,
+      "loss": 4.1293,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15331575778501e-05,
+      "loss": 4.1441,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152477163033958e-05,
+      "loss": 4.1438,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151638568282906e-05,
+      "loss": 4.1216,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150799973531853e-05,
+      "loss": 4.1258,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149961378780801e-05,
+      "loss": 4.1319,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149122784029749e-05,
+      "loss": 4.1244,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148284189278697e-05,
+      "loss": 4.1377,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147445594527645e-05,
+      "loss": 4.1165,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146606999776593e-05,
+      "loss": 4.1374,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145768405025541e-05,
+      "loss": 4.1228,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144929810274489e-05,
+      "loss": 4.1353,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144092853403811e-05,
+      "loss": 4.1218,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1432542586527587e-05,
+      "loss": 4.127,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1424156639017067e-05,
+      "loss": 4.1272,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1415770691506546e-05,
+      "loss": 4.1384,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1407384743996026e-05,
+      "loss": 4.1271,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1399031554092965e-05,
+      "loss": 4.1314,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1390645606582445e-05,
+      "loss": 4.1207,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382259659071925e-05,
+      "loss": 4.1363,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1373873711561405e-05,
+      "loss": 4.1281,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1365487764050885e-05,
+      "loss": 4.1329,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357101816540364e-05,
+      "loss": 4.1244,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1348715869029844e-05,
+      "loss": 4.135,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340329921519324e-05,
+      "loss": 4.1256,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1331943974008804e-05,
+      "loss": 4.1419,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132357440530202e-05,
+      "loss": 4.1192,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13151884577915e-05,
+      "loss": 4.1321,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130680251028098e-05,
+      "loss": 4.1254,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129841656277046e-05,
+      "loss": 4.1258,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129004699406367e-05,
+      "loss": 4.1247,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128167742535688e-05,
+      "loss": 4.1249,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127329147784636e-05,
+      "loss": 4.1286,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126490553033584e-05,
+      "loss": 4.1235,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125651958282532e-05,
+      "loss": 4.1336,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.133406639099121,
+      "eval_runtime": 297.6067,
+      "eval_samples_per_second": 1282.199,
+      "eval_steps_per_second": 40.07,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12481336353148e-05,
+      "loss": 4.122,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123974768780428e-05,
+      "loss": 4.1159,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123136174029376e-05,
+      "loss": 4.1214,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1222975792783245e-05,
+      "loss": 4.1221,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121458984527272e-05,
+      "loss": 4.132,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12062038977622e-05,
+      "loss": 4.1278,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119781795025168e-05,
+      "loss": 4.1202,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118943200274116e-05,
+      "loss": 4.1205,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118104605523064e-05,
+      "loss": 4.1314,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117266010772012e-05,
+      "loss": 4.1225,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11642741602096e-05,
+      "loss": 4.1202,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.1183,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114751864399229e-05,
+      "loss": 4.1218,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113913269648177e-05,
+      "loss": 4.117,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1130746748971247e-05,
+      "loss": 4.1201,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122360801460727e-05,
+      "loss": 4.1043,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.1124,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110558890643969e-05,
+      "loss": 4.1132,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.1091,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108883339022238e-05,
+      "loss": 4.1257,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108044744271186e-05,
+      "loss": 4.1253,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107206149520134e-05,
+      "loss": 4.118,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106367554769082e-05,
+      "loss": 4.1109,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10552896001803e-05,
+      "loss": 4.1252,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1046903652669775e-05,
+      "loss": 4.1148,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038517705159255e-05,
+      "loss": 4.118,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030131757648735e-05,
+      "loss": 4.1079,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021745810138215e-05,
+      "loss": 4.1195,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1013359862627695e-05,
+      "loss": 4.1203,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100497391511718e-05,
+      "loss": 4.1074,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099658796760666e-05,
+      "loss": 4.1153,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098821839889987e-05,
+      "loss": 4.1221,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097983245138935e-05,
+      "loss": 4.115,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097144650387883e-05,
+      "loss": 4.1274,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096306055636831e-05,
+      "loss": 4.1099,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095470736646525e-05,
+      "loss": 4.1183,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094632141895473e-05,
+      "loss": 4.1128,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093793547144421e-05,
+      "loss": 4.1025,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092954952393369e-05,
+      "loss": 4.108,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092116357642317e-05,
+      "loss": 4.1052,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0912794007716385e-05,
+      "loss": 4.1085,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904408060205865e-05,
+      "loss": 4.1221,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089603849149908e-05,
+      "loss": 4.112,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0887652543988554e-05,
+      "loss": 4.1121,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879266596478034e-05,
+      "loss": 4.1173,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0870880648967514e-05,
+      "loss": 4.1083,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0862494701456994e-05,
+      "loss": 4.1123,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0854108753946474e-05,
+      "loss": 4.1093,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0845722806435953e-05,
+      "loss": 4.0878,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0837336858925433e-05,
+      "loss": 4.123,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0828950911414913e-05,
+      "loss": 4.1095,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082058134270812e-05,
+      "loss": 4.1081,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08121953951976e-05,
+      "loss": 4.1068,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080380944768708e-05,
+      "loss": 4.1063,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07954398789803e-05,
+      "loss": 4.1001,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078705393146978e-05,
+      "loss": 4.1052,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077866798395926e-05,
+      "loss": 4.1083,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077028203644874e-05,
+      "loss": 4.11,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076191246774195e-05,
+      "loss": 4.1087,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075352652023143e-05,
+      "loss": 4.0988,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074514057272091e-05,
+      "loss": 4.1024,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073675462521039e-05,
+      "loss": 4.1157,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072836867769987e-05,
+      "loss": 4.0921,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071998273018935e-05,
+      "loss": 4.1009,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071159678267883e-05,
+      "loss": 4.1103,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070321083516831e-05,
+      "loss": 4.1081,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069482488765779e-05,
+      "loss": 4.101,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068643894014727e-05,
+      "loss": 4.1029,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067805299263675e-05,
+      "loss": 4.0929,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066968342392996e-05,
+      "loss": 4.1065,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066131385522317e-05,
+      "loss": 4.1168,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065292790771265e-05,
+      "loss": 4.1073,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064454196020213e-05,
+      "loss": 4.1128,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063615601269161e-05,
+      "loss": 4.1147,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0627770065181085e-05,
+      "loss": 4.1185,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0619384117670565e-05,
+      "loss": 4.1012,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0610998170160045e-05,
+      "loss": 4.1057,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0602612222649525e-05,
+      "loss": 4.1055,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594226275139005e-05,
+      "loss": 4.0889,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058585670643222e-05,
+      "loss": 4.1068,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05774707589217e-05,
+      "loss": 4.1011,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056908481141118e-05,
+      "loss": 4.1129,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056069886390066e-05,
+      "loss": 4.1197,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055231291639014e-05,
+      "loss": 4.0932,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054394334768335e-05,
+      "loss": 4.111,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053555740017283e-05,
+      "loss": 4.1007,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052717145266231e-05,
+      "loss": 4.1072,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051878550515179e-05,
+      "loss": 4.1011,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510415936445e-05,
+      "loss": 4.0998,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050202998893448e-05,
+      "loss": 4.0997,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049364404142396e-05,
+      "loss": 4.0998,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485258093913445e-05,
+      "loss": 4.1021,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0476872146402925e-05,
+      "loss": 4.0995,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468486198892405e-05,
+      "loss": 4.1088,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460100251381885e-05,
+      "loss": 4.1064,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451714303871365e-05,
+      "loss": 4.0957,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443344735164574e-05,
+      "loss": 4.0933,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434958787654054e-05,
+      "loss": 4.0974,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0426572840143534e-05,
+      "loss": 4.1023,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418186892633014e-05,
+      "loss": 4.1062,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0409800945122494e-05,
+      "loss": 4.0983,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04014313764157e-05,
+      "loss": 4.0953,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039304542890518e-05,
+      "loss": 4.0948,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03846758601984e-05,
+      "loss": 4.1113,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037628991268788e-05,
+      "loss": 4.0839,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036790396517736e-05,
+      "loss": 4.0993,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035951801766684e-05,
+      "loss": 4.0894,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035113207015632e-05,
+      "loss": 4.0983,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03427461226458e-05,
+      "loss": 4.089,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033436017513527e-05,
+      "loss": 4.1013,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032597422762475e-05,
+      "loss": 4.0874,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031758828011423e-05,
+      "loss": 4.0979,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030920233260371e-05,
+      "loss": 4.099,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030083276389692e-05,
+      "loss": 4.0902,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02924468163864e-05,
+      "loss": 4.0985,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028406086887588e-05,
+      "loss": 4.1143,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027567492136537e-05,
+      "loss": 4.1007,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026728897385485e-05,
+      "loss": 4.0994,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025890302634433e-05,
+      "loss": 4.0847,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250533457637536e-05,
+      "loss": 4.0987,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242147510127016e-05,
+      "loss": 4.097,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233761562616496e-05,
+      "loss": 4.0992,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225375615105976e-05,
+      "loss": 4.0829,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0216989667595456e-05,
+      "loss": 4.1008,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0208603720084936e-05,
+      "loss": 4.0935,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200217772574416e-05,
+      "loss": 4.0998,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191848203867625e-05,
+      "loss": 4.0869,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183462256357105e-05,
+      "loss": 4.0971,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175076308846585e-05,
+      "loss": 4.091,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166690361336065e-05,
+      "loss": 4.1074,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0158304413825545e-05,
+      "loss": 4.0955,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014993484511876e-05,
+      "loss": 4.0992,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014154889760824e-05,
+      "loss": 4.0853,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013316295009772e-05,
+      "loss": 4.1048,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01247770025872e-05,
+      "loss": 4.098,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011639105507668e-05,
+      "loss": 4.1012,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010800510756616e-05,
+      "loss": 4.0929,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0099619160055634e-05,
+      "loss": 4.0984,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0091233212545114e-05,
+      "loss": 4.0902,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0082847265034594e-05,
+      "loss": 4.1094,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007447769632781e-05,
+      "loss": 4.0879,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006609174881728e-05,
+      "loss": 4.1023,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00577221801105e-05,
+      "loss": 4.0956,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049336232599985e-05,
+      "loss": 4.0897,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0040950285089465e-05,
+      "loss": 4.0903,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003256433757894e-05,
+      "loss": 4.0933,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002417839006842e-05,
+      "loss": 4.0967,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00157924425579e-05,
+      "loss": 4.0927,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000740649504738e-05,
+      "loss": 4.0998,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.10874605178833,
+      "eval_runtime": 294.0942,
+      "eval_samples_per_second": 1297.513,
+      "eval_steps_per_second": 40.548,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999902054753686e-05,
+      "loss": 4.0861,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999063460002634e-05,
+      "loss": 4.0852,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998224865251582e-05,
+      "loss": 4.0867,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99738627050053e-05,
+      "loss": 4.0927,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996547675749478e-05,
+      "loss": 4.102,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995709080998426e-05,
+      "loss": 4.091,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994870486247374e-05,
+      "loss": 4.093,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994031891496322e-05,
+      "loss": 4.0876,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99319329674527e-05,
+      "loss": 4.0962,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992354701994218e-05,
+      "loss": 4.0936,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991516107243166e-05,
+      "loss": 4.0876,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990677512492114e-05,
+      "loss": 4.0841,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989838917741062e-05,
+      "loss": 4.097,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98900032299001e-05,
+      "loss": 4.0831,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988161728238958e-05,
+      "loss": 4.0831,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987323133487906e-05,
+      "loss": 4.078,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986484538736854e-05,
+      "loss": 4.0785,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985647581866175e-05,
+      "loss": 4.0852,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984808987115123e-05,
+      "loss": 4.0751,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839720302444436e-05,
+      "loss": 4.0939,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983133435493392e-05,
+      "loss": 4.0944,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98229484074234e-05,
+      "loss": 4.0918,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981456245991288e-05,
+      "loss": 4.0795,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980617651240236e-05,
+      "loss": 4.0888,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979779056489184e-05,
+      "loss": 4.0867,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978940461738132e-05,
+      "loss": 4.0851,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97810186698708e-05,
+      "loss": 4.0796,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977263272236028e-05,
+      "loss": 4.0897,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976424677484976e-05,
+      "loss": 4.0847,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9755860827339235e-05,
+      "loss": 4.0799,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747474879828715e-05,
+      "loss": 4.0807,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973910531112193e-05,
+      "loss": 4.0941,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973073574241514e-05,
+      "loss": 4.0903,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972234979490462e-05,
+      "loss": 4.0944,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97139638473941e-05,
+      "loss": 4.078,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970557789988359e-05,
+      "loss": 4.0846,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697208331176796e-05,
+      "loss": 4.0891,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9688822383666276e-05,
+      "loss": 4.0745,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680452814959485e-05,
+      "loss": 4.0759,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672066867448965e-05,
+      "loss": 4.0734,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663680919938445e-05,
+      "loss": 4.0789,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655294972427925e-05,
+      "loss": 4.0851,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9646909024917405e-05,
+      "loss": 4.087,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638523077406885e-05,
+      "loss": 4.0872,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963013712989636e-05,
+      "loss": 4.0809,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962175118238584e-05,
+      "loss": 4.0827,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613365234875325e-05,
+      "loss": 4.0779,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9604979287364805e-05,
+      "loss": 4.0819,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9596593339854285e-05,
+      "loss": 4.0539,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588207392343765e-05,
+      "loss": 4.0951,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9579821444833244e-05,
+      "loss": 4.0768,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571435497322724e-05,
+      "loss": 4.0853,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9563049549812204e-05,
+      "loss": 4.0759,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554663602301684e-05,
+      "loss": 4.0766,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9546294033594893e-05,
+      "loss": 4.0699,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953790808608437e-05,
+      "loss": 4.0744,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952952213857385e-05,
+      "loss": 4.0794,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952113619106333e-05,
+      "loss": 4.0777,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951276662235654e-05,
+      "loss": 4.0821,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950439705364976e-05,
+      "loss": 4.0731,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949601110613924e-05,
+      "loss": 4.0686,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948762515862872e-05,
+      "loss": 4.0895,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94792392111182e-05,
+      "loss": 4.0611,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947086964241141e-05,
+      "loss": 4.0691,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946248369490089e-05,
+      "loss": 4.0828,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945409774739037e-05,
+      "loss": 4.0838,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445728178683576e-05,
+      "loss": 4.0658,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437342231173056e-05,
+      "loss": 4.0794,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9428956283662536e-05,
+      "loss": 4.0596,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420570336152016e-05,
+      "loss": 4.0829,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941220076744523e-05,
+      "loss": 4.0832,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940381481993471e-05,
+      "loss": 4.0817,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939542887242419e-05,
+      "loss": 4.0797,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938704292491367e-05,
+      "loss": 4.0863,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937865697740315e-05,
+      "loss": 4.0921,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937027102989263e-05,
+      "loss": 4.0715,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936188508238211e-05,
+      "loss": 4.0781,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935351551367532e-05,
+      "loss": 4.0786,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93451295661648e-05,
+      "loss": 4.0546,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933674361865428e-05,
+      "loss": 4.0785,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932835767114376e-05,
+      "loss": 4.0755,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931997172363324e-05,
+      "loss": 4.0795,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931158577612272e-05,
+      "loss": 4.0966,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93031998286122e-05,
+      "loss": 4.067,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929483025990541e-05,
+      "loss": 4.0759,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286444312394896e-05,
+      "loss": 4.0742,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927805836488437e-05,
+      "loss": 4.0792,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926967241737385e-05,
+      "loss": 4.0701,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261319227470795e-05,
+      "loss": 4.0757,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9252933279960274e-05,
+      "loss": 4.0696,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9244547332449754e-05,
+      "loss": 4.0732,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236161384939234e-05,
+      "loss": 4.0693,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9227775437428714e-05,
+      "loss": 4.0749,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219389489918194e-05,
+      "loss": 4.08,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211019921211403e-05,
+      "loss": 4.073,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920263397370088e-05,
+      "loss": 4.0755,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919424802619036e-05,
+      "loss": 4.0636,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918586207867984e-05,
+      "loss": 4.0698,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917747613116932e-05,
+      "loss": 4.0733,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91690901836588e-05,
+      "loss": 4.0816,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916070423614828e-05,
+      "loss": 4.0707,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915233466744149e-05,
+      "loss": 4.0658,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914394871993097e-05,
+      "loss": 4.0727,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913556277242045e-05,
+      "loss": 4.0786,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912717682490993e-05,
+      "loss": 4.0591,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911879087739941e-05,
+      "loss": 4.0705,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911040492988889e-05,
+      "loss": 4.0553,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91020353611821e-05,
+      "loss": 4.0772,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909364941367159e-05,
+      "loss": 4.0612,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908526346616107e-05,
+      "loss": 4.0693,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907687751865055e-05,
+      "loss": 4.0621,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906849157114003e-05,
+      "loss": 4.0697,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906010562362951e-05,
+      "loss": 4.0734,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905171967611899e-05,
+      "loss": 4.0616,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90433501074122e-05,
+      "loss": 4.0691,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903496415990168e-05,
+      "loss": 4.0833,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026578212391157e-05,
+      "loss": 4.0784,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018192264880637e-05,
+      "loss": 4.0668,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9009806317370117e-05,
+      "loss": 4.0589,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9001420369859596e-05,
+      "loss": 4.0676,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993034422349076e-05,
+      "loss": 4.0751,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8984664853642286e-05,
+      "loss": 4.067,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897627890613177e-05,
+      "loss": 4.058,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896789295862125e-05,
+      "loss": 4.0729,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895950701111073e-05,
+      "loss": 4.0708,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8951121063600205e-05,
+      "loss": 4.0731,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8942735116089685e-05,
+      "loss": 4.0593,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89343655473829e-05,
+      "loss": 4.0695,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892597959987238e-05,
+      "loss": 4.0635,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917593652361854e-05,
+      "loss": 4.0787,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890922408365507e-05,
+      "loss": 4.0708,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890083813614455e-05,
+      "loss": 4.0695,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889245218863403e-05,
+      "loss": 4.0578,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888406624112351e-05,
+      "loss": 4.0808,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887568029361299e-05,
+      "loss": 4.068,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886729434610247e-05,
+      "loss": 4.0764,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885892477739568e-05,
+      "loss": 4.0671,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885053882988516e-05,
+      "loss": 4.0683,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884215288237464e-05,
+      "loss": 4.0667,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883376693486412e-05,
+      "loss": 4.0799,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88253809873536e-05,
+      "loss": 4.0615,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881699503984308e-05,
+      "loss": 4.0762,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880860909233256e-05,
+      "loss": 4.069,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880022314482204e-05,
+      "loss": 4.0629,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879183719731152e-05,
+      "loss": 4.0651,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783451249801e-05,
+      "loss": 4.0668,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877506530229048e-05,
+      "loss": 4.0678,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876667935477996e-05,
+      "loss": 4.0707,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875829340726944e-05,
+      "loss": 4.0754,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.089662551879883,
+      "eval_runtime": 301.0203,
+      "eval_samples_per_second": 1267.659,
+      "eval_steps_per_second": 39.615,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8749923838562654e-05,
+      "loss": 4.0566,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8741554269855863e-05,
+      "loss": 4.0591,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733168322345343e-05,
+      "loss": 4.0594,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8724782374834823e-05,
+      "loss": 4.068,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87163964273243e-05,
+      "loss": 4.0757,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870801047981378e-05,
+      "loss": 4.0629,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869962453230326e-05,
+      "loss": 4.0656,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869123858479274e-05,
+      "loss": 4.0621,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868286901608595e-05,
+      "loss": 4.0676,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867448306857543e-05,
+      "loss": 4.0724,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866609712106491e-05,
+      "loss": 4.0603,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865771117355439e-05,
+      "loss": 4.0579,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864932522604387e-05,
+      "loss": 4.0729,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864093927853335e-05,
+      "loss": 4.0559,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863255333102283e-05,
+      "loss": 4.0573,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862416738351231e-05,
+      "loss": 4.0519,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861579781480552e-05,
+      "loss": 4.0543,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607411867295e-05,
+      "loss": 4.0566,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859902591978448e-05,
+      "loss": 4.0513,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859063997227396e-05,
+      "loss": 4.0643,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858225402476344e-05,
+      "loss": 4.0672,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857386807725292e-05,
+      "loss": 4.0697,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85654821297424e-05,
+      "loss": 4.0546,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855709618223188e-05,
+      "loss": 4.0593,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8548742992328826e-05,
+      "loss": 4.0667,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8540357044818306e-05,
+      "loss": 4.0554,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8531971097307786e-05,
+      "loss": 4.0547,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8523585149797266e-05,
+      "loss": 4.0693,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8515199202286746e-05,
+      "loss": 4.0564,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8506813254776226e-05,
+      "loss": 4.0526,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8498443686069435e-05,
+      "loss": 4.0539,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490057738558915e-05,
+      "loss": 4.0649,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8481671791048395e-05,
+      "loss": 4.0653,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8473285843537875e-05,
+      "loss": 4.0733,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8464916274831084e-05,
+      "loss": 4.0511,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8456530327320564e-05,
+      "loss": 4.0598,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844814437981005e-05,
+      "loss": 4.0664,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843975843229953e-05,
+      "loss": 4.0422,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843137248478901e-05,
+      "loss": 4.0511,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842298653727849e-05,
+      "loss": 4.0506,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841460058976797e-05,
+      "loss": 4.0546,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840621464225745e-05,
+      "loss": 4.0609,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839784507355066e-05,
+      "loss": 4.0587,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838947550484387e-05,
+      "loss": 4.061,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838108955733335e-05,
+      "loss": 4.0552,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837270360982283e-05,
+      "loss": 4.057,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836431766231231e-05,
+      "loss": 4.0548,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835594809360552e-05,
+      "loss": 4.0556,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8347562146095004e-05,
+      "loss": 4.0281,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8339176198584484e-05,
+      "loss": 4.0683,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8330790251073964e-05,
+      "loss": 4.0504,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8322404303563444e-05,
+      "loss": 4.0609,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831403473485665e-05,
+      "loss": 4.0498,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830564878734613e-05,
+      "loss": 4.0535,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829726283983561e-05,
+      "loss": 4.0454,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828887689232509e-05,
+      "loss": 4.0479,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828049094481457e-05,
+      "loss": 4.0608,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272104997304046e-05,
+      "loss": 4.0464,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263719049793526e-05,
+      "loss": 4.0573,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255333102283006e-05,
+      "loss": 4.052,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824697991237996e-05,
+      "loss": 4.0378,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823859396486944e-05,
+      "loss": 4.0653,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823020801735892e-05,
+      "loss": 4.0371,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82218220698484e-05,
+      "loss": 4.0454,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821343612233788e-05,
+      "loss": 4.0602,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820505017482735e-05,
+      "loss": 4.0541,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819666422731683e-05,
+      "loss": 4.0404,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8188294658610047e-05,
+      "loss": 4.0557,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817990871109952e-05,
+      "loss": 4.0324,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8171522763589e-05,
+      "loss": 4.0587,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816313681607848e-05,
+      "loss": 4.0582,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8154767247371695e-05,
+      "loss": 4.0523,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8146381299861175e-05,
+      "loss": 4.0545,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8137995352350655e-05,
+      "loss": 4.0628,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8129609404840135e-05,
+      "loss": 4.0674,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812123983613335e-05,
+      "loss": 4.0464,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8112853888622824e-05,
+      "loss": 4.0514,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8104467941112304e-05,
+      "loss": 4.0587,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8096081993601784e-05,
+      "loss": 4.0278,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8087696046091264e-05,
+      "loss": 4.0521,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8079310098580744e-05,
+      "loss": 4.0539,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8070924151070224e-05,
+      "loss": 4.0548,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8062538203559704e-05,
+      "loss": 4.0708,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805416863485291e-05,
+      "loss": 4.0411,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804578268734239e-05,
+      "loss": 4.0509,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803739673983187e-05,
+      "loss": 4.0495,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802901079232136e-05,
+      "loss": 4.052,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802062484481084e-05,
+      "loss": 4.0505,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801225527610405e-05,
+      "loss": 4.0487,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800386932859353e-05,
+      "loss": 4.0421,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799548338108301e-05,
+      "loss": 4.0505,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798709743357249e-05,
+      "loss": 4.0465,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79787278648657e-05,
+      "loss": 4.0483,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797034191735518e-05,
+      "loss": 4.0571,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796197234864839e-05,
+      "loss": 4.0455,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953602779941596e-05,
+      "loss": 4.0511,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794521683243108e-05,
+      "loss": 4.0402,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793683088492056e-05,
+      "loss": 4.0435,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792844493741004e-05,
+      "loss": 4.0485,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792005898989952e-05,
+      "loss": 4.0547,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7911673042389e-05,
+      "loss": 4.0469,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790328709487848e-05,
+      "loss": 4.0422,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789490114736796e-05,
+      "loss": 4.0491,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788653157866117e-05,
+      "loss": 4.0504,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787814563115065e-05,
+      "loss": 4.0347,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786977606244386e-05,
+      "loss": 4.0466,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786139011493334e-05,
+      "loss": 4.0303,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785300416742282e-05,
+      "loss": 4.0521,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78446182199123e-05,
+      "loss": 4.0398,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783623227240178e-05,
+      "loss": 4.0434,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782784632489127e-05,
+      "loss": 4.0358,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781946037738075e-05,
+      "loss": 4.0516,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781107442987023e-05,
+      "loss": 4.0452,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780268848235971e-05,
+      "loss": 4.0395,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779430253484919e-05,
+      "loss": 4.0485,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778591658733866e-05,
+      "loss": 4.0568,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777753063982814e-05,
+      "loss": 4.055,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776914469231762e-05,
+      "loss": 4.0432,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776077512361083e-05,
+      "loss": 4.0353,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752405554904045e-05,
+      "loss": 4.0461,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744019607393525e-05,
+      "loss": 4.0516,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735633659883005e-05,
+      "loss": 4.0378,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7727247712372485e-05,
+      "loss": 4.0369,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7718861764861965e-05,
+      "loss": 4.0507,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7710475817351445e-05,
+      "loss": 4.0495,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7702089869840925e-05,
+      "loss": 4.048,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7693703922330405e-05,
+      "loss": 4.0361,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7685334353623614e-05,
+      "loss": 4.0455,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7676948406113094e-05,
+      "loss": 4.0399,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7668562458602574e-05,
+      "loss": 4.0552,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7660176511092054e-05,
+      "loss": 4.0411,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765180694238526e-05,
+      "loss": 4.0491,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764342099487474e-05,
+      "loss": 4.0304,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763503504736422e-05,
+      "loss": 4.0592,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76266490998537e-05,
+      "loss": 4.0385,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7618295909950654e-05,
+      "loss": 4.0567,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760990996244013e-05,
+      "loss": 4.0437,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760152401492961e-05,
+      "loss": 4.0478,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759313806741909e-05,
+      "loss": 4.0378,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758475211990857e-05,
+      "loss": 4.0592,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757636617239805e-05,
+      "loss": 4.0373,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756798022488753e-05,
+      "loss": 4.0501,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755959427737701e-05,
+      "loss": 4.0473,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7551224708670216e-05,
+      "loss": 4.038,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7542855139963426e-05,
+      "loss": 4.0443,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753446919245291e-05,
+      "loss": 4.0401,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752608324494239e-05,
+      "loss": 4.0438,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751769729743187e-05,
+      "loss": 4.0469,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750931134992135e-05,
+      "loss": 4.0505,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.073251247406006,
+      "eval_runtime": 289.5134,
+      "eval_samples_per_second": 1318.043,
+      "eval_steps_per_second": 41.19,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750092540241083e-05,
+      "loss": 4.0351,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749253945490031e-05,
+      "loss": 4.0358,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748415350738979e-05,
+      "loss": 4.0365,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747576755987927e-05,
+      "loss": 4.0432,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746738161236875e-05,
+      "loss": 4.0522,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745899566485823e-05,
+      "loss": 4.0442,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745060971734771e-05,
+      "loss": 4.0418,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744222376983719e-05,
+      "loss": 4.0396,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433837822326665e-05,
+      "loss": 4.0444,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425451874816145e-05,
+      "loss": 4.0486,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417065927305625e-05,
+      "loss": 4.037,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7408679979795105e-05,
+      "loss": 4.0384,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740029403228459e-05,
+      "loss": 4.0498,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739190808477407e-05,
+      "loss": 4.0333,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738352213726355e-05,
+      "loss": 4.0356,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737513618975303e-05,
+      "loss": 4.0251,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736675024224251e-05,
+      "loss": 4.0334,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735836429473199e-05,
+      "loss": 4.0371,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734997834722147e-05,
+      "loss": 4.0261,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734159239971095e-05,
+      "loss": 4.0409,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733322283100416e-05,
+      "loss": 4.0449,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732483688349364e-05,
+      "loss": 4.0441,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731648369359058e-05,
+      "loss": 4.0347,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730809774608006e-05,
+      "loss": 4.0385,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7299711798569545e-05,
+      "loss": 4.0444,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7291325851059025e-05,
+      "loss": 4.0314,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7282939903548505e-05,
+      "loss": 4.0266,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7274553956037985e-05,
+      "loss": 4.0467,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7266168008527465e-05,
+      "loss": 4.0336,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7257782061016945e-05,
+      "loss": 4.0335,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7249396113506425e-05,
+      "loss": 4.0301,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7241010165995905e-05,
+      "loss": 4.0419,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723262421848538e-05,
+      "loss": 4.0384,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722423827097486e-05,
+      "loss": 4.0538,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7215868702268074e-05,
+      "loss": 4.0319,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7207482754757554e-05,
+      "loss": 4.034,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719909680724703e-05,
+      "loss": 4.0483,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719072723854024e-05,
+      "loss": 4.0165,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718234129102973e-05,
+      "loss": 4.032,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717395534351921e-05,
+      "loss": 4.0305,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716556939600868e-05,
+      "loss": 4.0274,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715718344849816e-05,
+      "loss": 4.0409,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714879750098764e-05,
+      "loss": 4.0373,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714041155347712e-05,
+      "loss": 4.0405,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71320256059666e-05,
+      "loss": 4.0306,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712365603725981e-05,
+      "loss": 4.0367,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711527008974929e-05,
+      "loss": 4.0278,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71069005210425e-05,
+      "loss": 4.0387,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709851457353198e-05,
+      "loss": 4.0084,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709012862602147e-05,
+      "loss": 4.0414,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708174267851095e-05,
+      "loss": 4.0295,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707335673100043e-05,
+      "loss": 4.0398,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706497078348991e-05,
+      "loss": 4.0278,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705658483597939e-05,
+      "loss": 4.0307,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704819888846887e-05,
+      "loss": 4.0227,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7039829319762076e-05,
+      "loss": 4.0256,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7031443372251556e-05,
+      "loss": 4.0413,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7023057424741036e-05,
+      "loss": 4.0208,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014687856034245e-05,
+      "loss": 4.0363,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006318287327454e-05,
+      "loss": 4.0347,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6997932339816934e-05,
+      "loss": 4.0156,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698954639230642e-05,
+      "loss": 4.0428,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69811604447959e-05,
+      "loss": 4.0215,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697277449728538e-05,
+      "loss": 4.023,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696438854977486e-05,
+      "loss": 4.0356,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695600260226434e-05,
+      "loss": 4.0325,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694761665475382e-05,
+      "loss": 4.0205,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693924708604703e-05,
+      "loss": 4.0338,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930893896143975e-05,
+      "loss": 4.0142,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692250794863345e-05,
+      "loss": 4.036,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691412200112293e-05,
+      "loss": 4.0357,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690573605361241e-05,
+      "loss": 4.0329,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689735010610189e-05,
+      "loss": 4.0338,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688896415859137e-05,
+      "loss": 4.0395,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880594589884584e-05,
+      "loss": 4.0442,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872208642374064e-05,
+      "loss": 4.0248,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6863822694863544e-05,
+      "loss": 4.0315,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6855436747353024e-05,
+      "loss": 4.0321,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6847050799842504e-05,
+      "loss": 4.0115,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6838664852331984e-05,
+      "loss": 4.0276,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6830278904821464e-05,
+      "loss": 4.0358,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6821892957310944e-05,
+      "loss": 4.0302,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6813507009800424e-05,
+      "loss": 4.0508,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680513744109363e-05,
+      "loss": 4.0167,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679675149358311e-05,
+      "loss": 4.0298,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678838192487632e-05,
+      "loss": 4.0319,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677999597736581e-05,
+      "loss": 4.0302,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677161002985529e-05,
+      "loss": 4.0311,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676322408234477e-05,
+      "loss": 4.0287,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675483813483425e-05,
+      "loss": 4.0175,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674645218732373e-05,
+      "loss": 4.0291,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673808261861694e-05,
+      "loss": 4.0296,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672969667110642e-05,
+      "loss": 4.0239,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67213107235959e-05,
+      "loss": 4.0347,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671292477608538e-05,
+      "loss": 4.0308,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670453882857486e-05,
+      "loss": 4.0244,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669615288106434e-05,
+      "loss": 4.0205,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687783312357546e-05,
+      "loss": 4.0224,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6679397364847026e-05,
+      "loss": 4.0234,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671011417336506e-05,
+      "loss": 4.0356,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662625469825986e-05,
+      "loss": 4.0294,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66542559011192e-05,
+      "loss": 4.022,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664586995360868e-05,
+      "loss": 4.0269,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663748400609816e-05,
+      "loss": 4.0244,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629098058587635e-05,
+      "loss": 4.0206,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620712111077115e-05,
+      "loss": 4.0258,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6612326163566595e-05,
+      "loss": 4.0111,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6603940216056075e-05,
+      "loss": 4.0324,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595554268545555e-05,
+      "loss": 4.0231,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587184699838764e-05,
+      "loss": 4.0176,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6578798752328244e-05,
+      "loss": 4.0149,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657042918362146e-05,
+      "loss": 4.0276,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656204323611094e-05,
+      "loss": 4.0257,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655365728860042e-05,
+      "loss": 4.0177,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65452713410899e-05,
+      "loss": 4.027,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653688539357938e-05,
+      "loss": 4.0394,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652851582487259e-05,
+      "loss": 4.0317,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652012987736207e-05,
+      "loss": 4.0239,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651174392985155e-05,
+      "loss": 4.0161,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650335798234103e-05,
+      "loss": 4.0211,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649497203483051e-05,
+      "loss": 4.0349,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648658608731999e-05,
+      "loss": 4.0178,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647820013980947e-05,
+      "loss": 4.0172,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646981419229895e-05,
+      "loss": 4.0296,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461444623592164e-05,
+      "loss": 4.0267,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6453058676081644e-05,
+      "loss": 4.025,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6444672728571124e-05,
+      "loss": 4.0188,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643631953866806e-05,
+      "loss": 4.0234,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642793359115754e-05,
+      "loss": 4.0198,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641954764364702e-05,
+      "loss": 4.0349,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64111616961365e-05,
+      "loss": 4.022,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640277574862598e-05,
+      "loss": 4.0295,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639438980111546e-05,
+      "loss": 4.0133,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638600385360494e-05,
+      "loss": 4.0335,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637761790609442e-05,
+      "loss": 4.0247,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63692319585839e-05,
+      "loss": 4.0319,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636084601107338e-05,
+      "loss": 4.0276,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635246006356286e-05,
+      "loss": 4.0237,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634407411605235e-05,
+      "loss": 4.025,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633568816854182e-05,
+      "loss": 4.0363,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63273022210313e-05,
+      "loss": 4.0196,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631893265232452e-05,
+      "loss": 4.0261,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310546704814e-05,
+      "loss": 4.0296,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630216075730347e-05,
+      "loss": 4.0189,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629377480979295e-05,
+      "loss": 4.0221,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628538886228243e-05,
+      "loss": 4.0249,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627701929357564e-05,
+      "loss": 4.022,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626863334606512e-05,
+      "loss": 4.0279,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62602473985546e-05,
+      "loss": 4.0282,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0601935386657715,
+      "eval_runtime": 310.1606,
+      "eval_samples_per_second": 1230.301,
+      "eval_steps_per_second": 38.448,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251861451044087e-05,
+      "loss": 4.0138,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243475503533566e-05,
+      "loss": 4.019,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6235105934826776e-05,
+      "loss": 4.0193,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6226719987316256e-05,
+      "loss": 4.0234,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6218334039805736e-05,
+      "loss": 4.0323,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6209948092295215e-05,
+      "loss": 4.025,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6201562144784695e-05,
+      "loss": 4.0226,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6193192576077905e-05,
+      "loss": 4.0208,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6184823007371114e-05,
+      "loss": 4.0228,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176437059860594e-05,
+      "loss": 4.0246,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168051112350074e-05,
+      "loss": 4.0234,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6159665164839553e-05,
+      "loss": 4.0171,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615127921732904e-05,
+      "loss": 4.0289,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614289326981852e-05,
+      "loss": 4.0121,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6134507322308e-05,
+      "loss": 4.0178,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612612137479748e-05,
+      "loss": 4.0044,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611773542728696e-05,
+      "loss": 4.0175,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610936585858017e-05,
+      "loss": 4.0154,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610097991106965e-05,
+      "loss": 4.0079,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609259396355913e-05,
+      "loss": 4.0211,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608420801604861e-05,
+      "loss": 4.022,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607582206853809e-05,
+      "loss": 4.0224,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60674524998313e-05,
+      "loss": 4.0163,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605906655232078e-05,
+      "loss": 4.0183,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605068060481026e-05,
+      "loss": 4.0259,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604229465729974e-05,
+      "loss": 4.0173,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6033908709789225e-05,
+      "loss": 4.0032,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6025522762278705e-05,
+      "loss": 4.0278,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6017136814768185e-05,
+      "loss": 4.0152,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600875086725766e-05,
+      "loss": 4.012,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6000381298550874e-05,
+      "loss": 4.0167,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5991995351040354e-05,
+      "loss": 4.02,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5983609403529834e-05,
+      "loss": 4.0185,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597522345601931e-05,
+      "loss": 4.036,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596685388731252e-05,
+      "loss": 4.0133,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595848431860573e-05,
+      "loss": 4.0168,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595009837109521e-05,
+      "loss": 4.0296,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594171242358469e-05,
+      "loss": 3.9958,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593332647607418e-05,
+      "loss": 4.0128,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592495690736739e-05,
+      "loss": 4.0116,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591657095985687e-05,
+      "loss": 4.0079,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590818501234635e-05,
+      "loss": 4.0217,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589979906483583e-05,
+      "loss": 4.017,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589141311732531e-05,
+      "loss": 4.0193,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588302716981478e-05,
+      "loss": 4.0151,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587464122230426e-05,
+      "loss": 4.0186,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586625527479374e-05,
+      "loss": 4.0065,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585786932728322e-05,
+      "loss": 4.023,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584949975857643e-05,
+      "loss": 3.9892,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5841130189869645e-05,
+      "loss": 4.0179,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583274424235913e-05,
+      "loss": 4.0123,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582437467365234e-05,
+      "loss": 4.0217,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581598872614182e-05,
+      "loss": 4.0045,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58076027786313e-05,
+      "loss": 4.0134,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5799216831120774e-05,
+      "loss": 4.0061,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5790830883610254e-05,
+      "loss": 4.0059,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5782444936099734e-05,
+      "loss": 4.0265,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5774058988589214e-05,
+      "loss": 3.9996,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5765673041078694e-05,
+      "loss": 4.0179,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57573034723719e-05,
+      "loss": 4.0171,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574891752486138e-05,
+      "loss": 3.9924,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574053157735087e-05,
+      "loss": 4.0274,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573214562984035e-05,
+      "loss": 4.0024,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572375968232983e-05,
+      "loss": 4.0053,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571539011362304e-05,
+      "loss": 4.0145,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570700416611252e-05,
+      "loss": 4.0162,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5698618218602e-05,
+      "loss": 4.0009,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569023227109148e-05,
+      "loss": 4.0151,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568186270238469e-05,
+      "loss": 3.9946,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567347675487417e-05,
+      "loss": 4.0162,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566509080736365e-05,
+      "loss": 4.0205,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565670485985313e-05,
+      "loss": 4.0154,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564831891234261e-05,
+      "loss": 4.0143,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563993296483209e-05,
+      "loss": 4.018,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5631563396125303e-05,
+      "loss": 4.0296,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562319382741851e-05,
+      "loss": 4.0062,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561480787990799e-05,
+      "loss": 4.0165,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560642193239747e-05,
+      "loss": 4.0083,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559803598488695e-05,
+      "loss": 4.0,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558965003737643e-05,
+      "loss": 4.0067,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558126408986591e-05,
+      "loss": 4.017,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557287814235539e-05,
+      "loss": 4.0137,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556449219484487e-05,
+      "loss": 4.0354,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555610624733435e-05,
+      "loss": 3.9957,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554772029982383e-05,
+      "loss": 4.0129,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553933435231331e-05,
+      "loss": 4.0137,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553096478360652e-05,
+      "loss": 4.0109,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522578836096e-05,
+      "loss": 4.0112,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551419288858549e-05,
+      "loss": 4.0091,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550580694107497e-05,
+      "loss": 4.006,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549742099356444e-05,
+      "loss": 4.0091,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548903504605392e-05,
+      "loss": 4.0128,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54806490985434e-05,
+      "loss": 4.0045,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547227952983661e-05,
+      "loss": 4.0171,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546389358232609e-05,
+      "loss": 4.0134,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545550763481557e-05,
+      "loss": 4.0056,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544712168730505e-05,
+      "loss": 4.0052,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543873573979453e-05,
+      "loss": 4.007,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543034979228401e-05,
+      "loss": 4.0061,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542196384477349e-05,
+      "loss": 4.0157,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541357789726297e-05,
+      "loss": 4.0111,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5405208328556186e-05,
+      "loss": 4.0001,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5396822381045665e-05,
+      "loss": 4.0118,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5388436433535145e-05,
+      "loss": 4.0078,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5380066864828355e-05,
+      "loss": 4.0029,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5371680917317835e-05,
+      "loss": 4.0038,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5363294969807314e-05,
+      "loss": 3.9951,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5354909022296794e-05,
+      "loss": 4.011,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5346523074786274e-05,
+      "loss": 4.0073,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5338153506079483e-05,
+      "loss": 3.9979,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532978393737269e-05,
+      "loss": 3.9959,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532139798986218e-05,
+      "loss": 4.0097,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531301204235166e-05,
+      "loss": 4.0089,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530462609484114e-05,
+      "loss": 3.9974,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529624014733062e-05,
+      "loss": 4.0101,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52878541998201e-05,
+      "loss": 4.0211,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527948463111331e-05,
+      "loss": 4.0111,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527109868360279e-05,
+      "loss": 4.0059,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526271273609227e-05,
+      "loss": 4.0001,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525434316738548e-05,
+      "loss": 4.0053,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524595721987496e-05,
+      "loss": 4.0167,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523757127236444e-05,
+      "loss": 4.0008,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522918532485392e-05,
+      "loss": 3.9984,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52207993773434e-05,
+      "loss": 4.0084,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521241342983288e-05,
+      "loss": 4.0073,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5204027482322364e-05,
+      "loss": 4.0091,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5195641534811844e-05,
+      "loss": 4.002,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5187255587301324e-05,
+      "loss": 4.0021,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51788696397908e-05,
+      "loss": 4.0028,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517048369228028e-05,
+      "loss": 4.0173,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516209774476976e-05,
+      "loss": 4.0046,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515371179725924e-05,
+      "loss": 4.0105,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514532584974872e-05,
+      "loss": 3.9969,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5136956281041926e-05,
+      "loss": 4.0114,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128570333531406e-05,
+      "loss": 4.0091,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5120184386020886e-05,
+      "loss": 4.0102,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111798438510366e-05,
+      "loss": 4.0125,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103412490999846e-05,
+      "loss": 4.005,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095026543489326e-05,
+      "loss": 4.0078,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508664059597881e-05,
+      "loss": 4.0189,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507825464846829e-05,
+      "loss": 4.0034,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50698850797615e-05,
+      "loss": 4.0069,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506151551105471e-05,
+      "loss": 4.0137,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505312956354419e-05,
+      "loss": 4.0011,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504474361603367e-05,
+      "loss": 4.0059,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503637404732688e-05,
+      "loss": 4.0102,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502798809981636e-05,
+      "loss": 4.0005,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501960215230584e-05,
+      "loss": 4.0072,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501121620479532e-05,
+      "loss": 4.0115,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.048716068267822,
+      "eval_runtime": 308.7765,
+      "eval_samples_per_second": 1235.816,
+      "eval_steps_per_second": 38.62,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.50028302572848e-05,
+      "loss": 4.0074,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499444430977428e-05,
+      "loss": 3.9963,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4986058362263766e-05,
+      "loss": 3.9994,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4977672414753246e-05,
+      "loss": 4.0081,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4969286467242726e-05,
+      "loss": 4.0072,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4960900519732206e-05,
+      "loss": 4.0091,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952514572221686e-05,
+      "loss": 4.0084,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494412862471116e-05,
+      "loss": 4.0022,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493574267720064e-05,
+      "loss": 4.0027,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492735672969012e-05,
+      "loss": 4.0067,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49189707821796e-05,
+      "loss": 4.0076,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491058483466908e-05,
+      "loss": 3.9962,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490221526596229e-05,
+      "loss": 4.0148,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489382931845177e-05,
+      "loss": 3.9939,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488544337094125e-05,
+      "loss": 3.9967,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4877057423430734e-05,
+      "loss": 3.9904,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4868671475920214e-05,
+      "loss": 4.0,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860285528409694e-05,
+      "loss": 3.9965,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4851899580899174e-05,
+      "loss": 3.992,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4843530012192383e-05,
+      "loss": 3.9986,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483516044348559e-05,
+      "loss": 4.0104,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482677449597507e-05,
+      "loss": 4.005,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481838854846455e-05,
+      "loss": 3.9975,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481000260095403e-05,
+      "loss": 4.0057,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480161665344351e-05,
+      "loss": 4.0076,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479323070593299e-05,
+      "loss": 3.9962,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478484475842247e-05,
+      "loss": 3.9915,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477645881091195e-05,
+      "loss": 4.0083,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476807286340143e-05,
+      "loss": 4.0014,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475968691589092e-05,
+      "loss": 3.9925,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47513009683804e-05,
+      "loss": 3.9976,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474291502086988e-05,
+      "loss": 3.9998,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473454545216309e-05,
+      "loss": 4.005,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472615950465257e-05,
+      "loss": 4.0196,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471777355714205e-05,
+      "loss": 3.9921,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470938760963152e-05,
+      "loss": 4.0052,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470101804092474e-05,
+      "loss": 4.0102,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469263209341422e-05,
+      "loss": 3.9784,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46842461459037e-05,
+      "loss": 3.9974,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467586019839317e-05,
+      "loss": 3.9964,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466747425088266e-05,
+      "loss": 3.9861,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465908830337214e-05,
+      "loss": 4.0106,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465070235586162e-05,
+      "loss": 3.997,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4642316408351097e-05,
+      "loss": 4.0044,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463396321844804e-05,
+      "loss": 3.9959,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462557727093752e-05,
+      "loss": 4.0023,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4617191323426995e-05,
+      "loss": 3.9885,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460882175472021e-05,
+      "loss": 4.0064,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460043580720969e-05,
+      "loss": 3.9757,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459204985969917e-05,
+      "loss": 3.999,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4583663912188644e-05,
+      "loss": 3.996,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575277964678124e-05,
+      "loss": 4.0078,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456690839597134e-05,
+      "loss": 3.9871,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455852244846082e-05,
+      "loss": 3.9943,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45501365009503e-05,
+      "loss": 3.9932,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454175055343978e-05,
+      "loss": 3.9889,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453336460592926e-05,
+      "loss": 4.0107,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452497865841874e-05,
+      "loss": 3.9781,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451659271090822e-05,
+      "loss": 4.0046,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45082067633977e-05,
+      "loss": 4.002,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449982081588718e-05,
+      "loss": 3.9764,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449145124718039e-05,
+      "loss": 4.008,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448306529966987e-05,
+      "loss": 3.9843,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447467935215935e-05,
+      "loss": 3.9872,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446629340464883e-05,
+      "loss": 3.9987,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445790745713831e-05,
+      "loss": 4.0005,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444952150962779e-05,
+      "loss": 3.9868,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4441135562117275e-05,
+      "loss": 3.9994,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4432749614606755e-05,
+      "loss": 3.9774,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4424363667096235e-05,
+      "loss": 3.9957,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441601047719317e-05,
+      "loss": 4.0051,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440762452968265e-05,
+      "loss": 3.9982,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439923858217213e-05,
+      "loss": 3.9999,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439085263466161e-05,
+      "loss": 3.9999,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438248306595482e-05,
+      "loss": 4.0104,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43740971184443e-05,
+      "loss": 3.9911,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436571117093378e-05,
+      "loss": 4.0005,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435732522342326e-05,
+      "loss": 3.9911,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434893927591274e-05,
+      "loss": 3.98,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434056970720596e-05,
+      "loss": 3.9923,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433218375969544e-05,
+      "loss": 4.0006,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432379781218492e-05,
+      "loss": 3.9997,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43154118646744e-05,
+      "loss": 4.0166,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430702591716388e-05,
+      "loss": 3.9875,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429863996965336e-05,
+      "loss": 3.9923,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429025402214283e-05,
+      "loss": 3.9939,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281900832239776e-05,
+      "loss": 4.0009,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4273514884729256e-05,
+      "loss": 3.992,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265128937218735e-05,
+      "loss": 3.9914,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4256742989708215e-05,
+      "loss": 3.9898,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4248357042197695e-05,
+      "loss": 3.9914,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423997109468718e-05,
+      "loss": 3.9956,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4231585147176655e-05,
+      "loss": 3.989,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4223199199666135e-05,
+      "loss": 4.0013,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4214813252155615e-05,
+      "loss": 3.998,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420644368344883e-05,
+      "loss": 3.9896,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4198057735938304e-05,
+      "loss": 3.987,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4189671788427784e-05,
+      "loss": 3.9877,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181285840917264e-05,
+      "loss": 3.9907,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4172899893406744e-05,
+      "loss": 4.0038,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164513945896224e-05,
+      "loss": 3.9926,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156127998385704e-05,
+      "loss": 3.9849,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414775842967892e-05,
+      "loss": 3.9948,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41393724821684e-05,
+      "loss": 3.9909,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413098653465788e-05,
+      "loss": 3.9922,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412261696595109e-05,
+      "loss": 3.9912,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411423101844057e-05,
+      "loss": 3.9777,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410584507093005e-05,
+      "loss": 3.9948,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409745912341953e-05,
+      "loss": 3.9896,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408907317590901e-05,
+      "loss": 3.9862,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408068722839849e-05,
+      "loss": 3.9779,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407230128088797e-05,
+      "loss": 3.993,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406391533337745e-05,
+      "loss": 3.9962,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405552938586693e-05,
+      "loss": 3.9815,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047176195963874e-05,
+      "loss": 3.9953,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4038790248453354e-05,
+      "loss": 4.0023,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4030404300942833e-05,
+      "loss": 4.0016,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4022018353432313e-05,
+      "loss": 3.9918,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401363240592179e-05,
+      "loss": 3.9836,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400524645841127e-05,
+      "loss": 3.9897,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399686051090075e-05,
+      "loss": 4.0028,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398847456339023e-05,
+      "loss": 3.9833,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398010499468344e-05,
+      "loss": 3.988,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397173542597665e-05,
+      "loss": 3.9864,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396334947846613e-05,
+      "loss": 3.9957,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395496353095561e-05,
+      "loss": 3.9937,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394657758344509e-05,
+      "loss": 3.9891,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393819163593457e-05,
+      "loss": 3.9852,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392980568842405e-05,
+      "loss": 3.9873,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392141974091354e-05,
+      "loss": 4.005,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391303379340302e-05,
+      "loss": 3.9888,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390464784589249e-05,
+      "loss": 3.9973,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389627827718571e-05,
+      "loss": 3.9772,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388789232967519e-05,
+      "loss": 3.9991,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387950638216467e-05,
+      "loss": 3.9935,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387112043465414e-05,
+      "loss": 3.9938,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386273448714362e-05,
+      "loss": 3.999,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38543485396331e-05,
+      "loss": 3.9852,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384596259212258e-05,
+      "loss": 3.9973,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383759302341579e-05,
+      "loss": 4.003,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3829207075905276e-05,
+      "loss": 3.9876,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3820821128394756e-05,
+      "loss": 3.9903,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3812435180884236e-05,
+      "loss": 3.9969,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3804049233373716e-05,
+      "loss": 3.9821,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3795679664666925e-05,
+      "loss": 3.9886,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378731009596014e-05,
+      "loss": 3.9975,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3778924148449614e-05,
+      "loss": 3.9861,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3770538200939094e-05,
+      "loss": 3.994,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3762152253428574e-05,
+      "loss": 3.9943,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.039572715759277,
+      "eval_runtime": 309.2409,
+      "eval_samples_per_second": 1233.96,
+      "eval_steps_per_second": 38.562,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3753766305918054e-05,
+      "loss": 3.9855,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374539673721126e-05,
+      "loss": 3.9798,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373701078970074e-05,
+      "loss": 3.9851,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372862484219023e-05,
+      "loss": 3.9925,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372023889467971e-05,
+      "loss": 3.994,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371185294716919e-05,
+      "loss": 3.9971,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370346699965867e-05,
+      "loss": 3.9869,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369508105214815e-05,
+      "loss": 3.9901,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368669510463763e-05,
+      "loss": 3.9862,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367832553593084e-05,
+      "loss": 3.9974,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366993958842032e-05,
+      "loss": 3.9874,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36615536409098e-05,
+      "loss": 3.9831,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365316769339928e-05,
+      "loss": 3.9977,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364479812469249e-05,
+      "loss": 3.9766,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363641217718197e-05,
+      "loss": 3.9838,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362802622967145e-05,
+      "loss": 3.9759,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361964028216093e-05,
+      "loss": 3.9843,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3611254334650414e-05,
+      "loss": 3.9817,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3602868387139894e-05,
+      "loss": 3.9778,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3594482439629374e-05,
+      "loss": 3.9778,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3586096492118854e-05,
+      "loss": 3.998,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357771054460833e-05,
+      "loss": 3.9884,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356934097590154e-05,
+      "loss": 3.9821,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356095502839102e-05,
+      "loss": 3.9904,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35525690808805e-05,
+      "loss": 3.9947,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3544183133369976e-05,
+      "loss": 3.9787,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353581356466319e-05,
+      "loss": 3.9765,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35274439959564e-05,
+      "loss": 3.9928,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351905804844588e-05,
+      "loss": 3.9822,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351067210093537e-05,
+      "loss": 3.9794,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350228615342485e-05,
+      "loss": 3.9838,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349390020591433e-05,
+      "loss": 3.9798,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34855142584038e-05,
+      "loss": 3.9947,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347712831089328e-05,
+      "loss": 3.9998,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346874236338276e-05,
+      "loss": 3.9818,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346037279467597e-05,
+      "loss": 3.9883,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345198684716545e-05,
+      "loss": 3.9964,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344360089965493e-05,
+      "loss": 3.9645,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343521495214441e-05,
+      "loss": 3.9881,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3426861762241354e-05,
+      "loss": 3.9752,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3418475814730834e-05,
+      "loss": 3.9724,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3410089867220314e-05,
+      "loss": 3.9915,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34017039197098e-05,
+      "loss": 3.989,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3393317972199274e-05,
+      "loss": 3.9857,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3384932024688754e-05,
+      "loss": 3.9798,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3376546077178234e-05,
+      "loss": 3.9842,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3368160129667714e-05,
+      "loss": 3.9785,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3359774182157194e-05,
+      "loss": 3.9891,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3351388234646674e-05,
+      "loss": 3.9632,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3343002287136154e-05,
+      "loss": 3.9805,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3334616339625634e-05,
+      "loss": 3.984,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3326230392115114e-05,
+      "loss": 3.9907,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3317844444604594e-05,
+      "loss": 3.9732,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309458497094074e-05,
+      "loss": 3.9792,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301072549583554e-05,
+      "loss": 3.9744,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329270298087677e-05,
+      "loss": 3.9761,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328431703336625e-05,
+      "loss": 3.9946,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327594746465946e-05,
+      "loss": 3.959,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326756151714894e-05,
+      "loss": 3.9903,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325917556963842e-05,
+      "loss": 3.9865,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32507896221279e-05,
+      "loss": 3.9636,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324240367461738e-05,
+      "loss": 3.995,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323403410591059e-05,
+      "loss": 3.9653,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322564815840007e-05,
+      "loss": 3.9715,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321726221088955e-05,
+      "loss": 3.9814,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320887626337903e-05,
+      "loss": 3.9864,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320049031586851e-05,
+      "loss": 3.9667,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319212074716172e-05,
+      "loss": 3.9903,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31837347996512e-05,
+      "loss": 3.9614,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317534885214068e-05,
+      "loss": 3.9768,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316696290463016e-05,
+      "loss": 3.9907,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3158576957119636e-05,
+      "loss": 3.9831,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3150191009609116e-05,
+      "loss": 3.9833,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3141805062098596e-05,
+      "loss": 3.9883,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133435493391805e-05,
+      "loss": 3.9971,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312506592468502e-05,
+      "loss": 3.9779,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31166799771745e-05,
+      "loss": 3.9871,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310829402966398e-05,
+      "loss": 3.9719,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309990808215346e-05,
+      "loss": 3.9686,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309152213464294e-05,
+      "loss": 3.9783,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308313618713242e-05,
+      "loss": 3.9818,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30747502396219e-05,
+      "loss": 3.9845,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306636429211138e-05,
+      "loss": 4.0009,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305797834460086e-05,
+      "loss": 3.9785,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049625154697806e-05,
+      "loss": 3.9742,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304123920718728e-05,
+      "loss": 3.9767,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303285325967676e-05,
+      "loss": 3.9879,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302446731216624e-05,
+      "loss": 3.9759,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301608136465572e-05,
+      "loss": 3.977,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30076954171452e-05,
+      "loss": 3.9777,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299930946963468e-05,
+      "loss": 3.9751,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299092352212416e-05,
+      "loss": 3.9813,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2982537574613646e-05,
+      "loss": 3.9745,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2974151627103126e-05,
+      "loss": 3.9855,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2965765679592606e-05,
+      "loss": 3.9847,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2957379732082085e-05,
+      "loss": 3.9738,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2948993784571565e-05,
+      "loss": 3.9748,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2940640594668504e-05,
+      "loss": 3.9731,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2932254647157984e-05,
+      "loss": 3.9762,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2923868699647464e-05,
+      "loss": 3.9887,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2915482752136944e-05,
+      "loss": 3.9819,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2907096804626423e-05,
+      "loss": 3.9671,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2898710857115903e-05,
+      "loss": 3.9838,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890324909605383e-05,
+      "loss": 3.9769,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288195534089859e-05,
+      "loss": 3.9726,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287356939338808e-05,
+      "loss": 3.9761,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286518344587756e-05,
+      "loss": 3.9664,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285679749836704e-05,
+      "loss": 3.9796,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284841155085652e-05,
+      "loss": 3.9768,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284004198214973e-05,
+      "loss": 3.9691,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283165603463921e-05,
+      "loss": 3.9657,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282327008712869e-05,
+      "loss": 3.9777,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281488413961817e-05,
+      "loss": 3.982,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280649819210764e-05,
+      "loss": 3.9709,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279812862340086e-05,
+      "loss": 3.9777,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278974267589034e-05,
+      "loss": 3.9844,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278135672837982e-05,
+      "loss": 3.9895,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27729707808693e-05,
+      "loss": 3.9806,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276460121216251e-05,
+      "loss": 3.9675,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275621526465199e-05,
+      "loss": 3.9761,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2747829317141466e-05,
+      "loss": 3.9864,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2739443369630946e-05,
+      "loss": 3.9689,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2731057422120426e-05,
+      "loss": 3.974,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2722671474609906e-05,
+      "loss": 3.9741,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2714285527099386e-05,
+      "loss": 3.9771,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2705899579588866e-05,
+      "loss": 3.9761,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2697530010882075e-05,
+      "loss": 3.9807,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268916044217529e-05,
+      "loss": 3.9668,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268077449466477e-05,
+      "loss": 3.9736,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267238854715425e-05,
+      "loss": 3.9906,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266400259964373e-05,
+      "loss": 3.9787,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265561665213321e-05,
+      "loss": 3.9791,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264724708342642e-05,
+      "loss": 3.9671,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26388611359159e-05,
+      "loss": 3.9835,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263047518840538e-05,
+      "loss": 3.9811,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262208924089486e-05,
+      "loss": 3.9814,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261370329338434e-05,
+      "loss": 3.9783,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260533372467755e-05,
+      "loss": 3.9743,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259694777716703e-05,
+      "loss": 3.986,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258856182965651e-05,
+      "loss": 3.9838,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258017588214599e-05,
+      "loss": 3.9812,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257178993463547e-05,
+      "loss": 3.9717,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563403987124955e-05,
+      "loss": 3.9856,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2555018039614435e-05,
+      "loss": 3.974,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2546632092103915e-05,
+      "loss": 3.9716,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253827890220085e-05,
+      "loss": 3.9873,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252989295469033e-05,
+      "loss": 3.969,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252150700717981e-05,
+      "loss": 3.9833,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251312105966929e-05,
+      "loss": 3.9766,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.031257152557373,
+      "eval_runtime": 311.3356,
+      "eval_samples_per_second": 1225.658,
+      "eval_steps_per_second": 38.303,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.25047514909625e-05,
+      "loss": 3.968,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249636554345198e-05,
+      "loss": 3.9694,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248797959594146e-05,
+      "loss": 3.9733,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247959364843094e-05,
+      "loss": 3.977,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247120770092042e-05,
+      "loss": 3.9822,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246282175340991e-05,
+      "loss": 3.9797,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245443580589939e-05,
+      "loss": 3.974,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244604985838887e-05,
+      "loss": 3.9749,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243766391087835e-05,
+      "loss": 3.9735,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242927796336783e-05,
+      "loss": 3.9842,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24208920158573e-05,
+      "loss": 3.9741,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241250606834678e-05,
+      "loss": 3.9731,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240413649964e-05,
+      "loss": 3.9845,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239575055212948e-05,
+      "loss": 3.9638,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238736460461895e-05,
+      "loss": 3.9685,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237897865710843e-05,
+      "loss": 3.9656,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370609088401647e-05,
+      "loss": 3.9721,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2362223140891127e-05,
+      "loss": 3.9647,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2353837193380606e-05,
+      "loss": 3.965,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2345451245870086e-05,
+      "loss": 3.9689,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2337065298359566e-05,
+      "loss": 3.9799,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2328679350849046e-05,
+      "loss": 3.977,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2320293403338526e-05,
+      "loss": 3.9658,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2311907455828006e-05,
+      "loss": 3.9792,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2303521508317486e-05,
+      "loss": 3.9813,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2295135560806966e-05,
+      "loss": 3.966,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2286749613296446e-05,
+      "loss": 3.9591,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2278363665785926e-05,
+      "loss": 3.9799,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2269977718275406e-05,
+      "loss": 3.9643,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2261608149568615e-05,
+      "loss": 3.9697,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2253222202058095e-05,
+      "loss": 3.9697,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2244836254547575e-05,
+      "loss": 3.9698,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223646668584079e-05,
+      "loss": 3.9785,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222808073833027e-05,
+      "loss": 3.9854,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221969479081975e-05,
+      "loss": 3.9705,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221130884330923e-05,
+      "loss": 3.9729,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220292289579871e-05,
+      "loss": 3.9804,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219453694828819e-05,
+      "loss": 3.9549,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2186151000777664e-05,
+      "loss": 3.9749,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217778143207088e-05,
+      "loss": 3.9605,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216939548456036e-05,
+      "loss": 3.9627,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216100953704984e-05,
+      "loss": 3.9738,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215262358953931e-05,
+      "loss": 3.9797,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214423764202879e-05,
+      "loss": 3.9679,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213585169451828e-05,
+      "loss": 3.9693,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212746574700776e-05,
+      "loss": 3.97,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211907979949724e-05,
+      "loss": 3.962,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2110726609594184e-05,
+      "loss": 3.9765,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2102340662083664e-05,
+      "loss": 3.9528,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209395471457314e-05,
+      "loss": 3.9665,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208556876706262e-05,
+      "loss": 3.9715,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20771828195521e-05,
+      "loss": 3.9799,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206879687204158e-05,
+      "loss": 3.9592,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206041092453106e-05,
+      "loss": 3.9619,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2052041355824267e-05,
+      "loss": 3.9613,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2043655408313746e-05,
+      "loss": 3.9603,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203526946080323e-05,
+      "loss": 3.9846,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202688351329271e-05,
+      "loss": 3.9521,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201849756578219e-05,
+      "loss": 3.9755,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201011161827167e-05,
+      "loss": 3.9727,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200172567076115e-05,
+      "loss": 3.9558,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199333972325063e-05,
+      "loss": 3.9766,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198495377574011e-05,
+      "loss": 3.9555,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197658420703332e-05,
+      "loss": 3.9588,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19681982595228e-05,
+      "loss": 3.9683,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195981231201228e-05,
+      "loss": 3.9755,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195144274330549e-05,
+      "loss": 3.9553,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194305679579497e-05,
+      "loss": 3.9766,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193467084828445e-05,
+      "loss": 3.9481,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192628490077393e-05,
+      "loss": 3.9641,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191789895326342e-05,
+      "loss": 3.9799,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19095130057529e-05,
+      "loss": 3.9725,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190114343704611e-05,
+      "loss": 3.9683,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189275748953559e-05,
+      "loss": 3.9748,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1884371542025067e-05,
+      "loss": 3.9835,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1875985594514547e-05,
+      "loss": 3.9657,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1867616025807756e-05,
+      "loss": 3.9794,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1859230078297236e-05,
+      "loss": 3.9584,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1850844130786716e-05,
+      "loss": 3.9563,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1842458183276196e-05,
+      "loss": 3.9623,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1834072235765675e-05,
+      "loss": 3.9693,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1825686288255155e-05,
+      "loss": 3.9702,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1817300340744635e-05,
+      "loss": 3.9902,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180893077203785e-05,
+      "loss": 3.9691,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1800544824527324e-05,
+      "loss": 3.9581,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1792158877016804e-05,
+      "loss": 3.9678,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1783772929506284e-05,
+      "loss": 3.9719,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1775386981995764e-05,
+      "loss": 3.9638,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1767001034485244e-05,
+      "loss": 3.9639,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758615086974724e-05,
+      "loss": 3.967,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175024551826793e-05,
+      "loss": 3.9611,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174185957075741e-05,
+      "loss": 3.9704,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173347362324689e-05,
+      "loss": 3.9594,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172508767573637e-05,
+      "loss": 3.9729,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171670172822585e-05,
+      "loss": 3.9734,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170831578071533e-05,
+      "loss": 3.9642,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169992983320482e-05,
+      "loss": 3.9586,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16915438856943e-05,
+      "loss": 3.962,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168319069579124e-05,
+      "loss": 3.9611,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167480474828072e-05,
+      "loss": 3.9752,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16664188007702e-05,
+      "loss": 3.9687,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165803285325968e-05,
+      "loss": 3.9569,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164964690574916e-05,
+      "loss": 3.9675,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164126095823864e-05,
+      "loss": 3.9624,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163287501072812e-05,
+      "loss": 3.9653,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16244890632176e-05,
+      "loss": 3.9641,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161613587331454e-05,
+      "loss": 3.9577,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160774992580402e-05,
+      "loss": 3.965,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15993639782935e-05,
+      "loss": 3.964,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159097803078298e-05,
+      "loss": 3.9557,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158259208327246e-05,
+      "loss": 3.9533,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157420613576194e-05,
+      "loss": 3.9606,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156582018825142e-05,
+      "loss": 3.9733,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155745061954463e-05,
+      "loss": 3.9587,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154906467203411e-05,
+      "loss": 3.965,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154067872452359e-05,
+      "loss": 3.9654,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153229277701307e-05,
+      "loss": 3.9826,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152390682950255e-05,
+      "loss": 3.9658,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151552088199203e-05,
+      "loss": 3.9541,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150713493448151e-05,
+      "loss": 3.9643,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149874898697099e-05,
+      "loss": 3.9692,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490395797067936e-05,
+      "loss": 3.9567,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1482009849557416e-05,
+      "loss": 3.962,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1473623902046896e-05,
+      "loss": 3.9624,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1465237954536376e-05,
+      "loss": 3.9657,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1456852007025856e-05,
+      "loss": 3.958,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1448466059515336e-05,
+      "loss": 3.967,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144008011200481e-05,
+      "loss": 3.9598,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143169416449429e-05,
+      "loss": 3.9637,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1423324595787505e-05,
+      "loss": 3.9744,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141493864827698e-05,
+      "loss": 3.9677,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1406552700766465e-05,
+      "loss": 3.9649,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139818313205968e-05,
+      "loss": 3.957,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138979718454916e-05,
+      "loss": 3.9696,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1381411237038634e-05,
+      "loss": 3.9687,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1373025289528114e-05,
+      "loss": 3.9681,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1364639342017594e-05,
+      "loss": 3.9707,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1356253394507074e-05,
+      "loss": 3.9601,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134788382580028e-05,
+      "loss": 3.9753,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133949787828976e-05,
+      "loss": 3.9706,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133111193077924e-05,
+      "loss": 3.966,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132272598326872e-05,
+      "loss": 3.9619,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13143400357582e-05,
+      "loss": 3.9712,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130597046705142e-05,
+      "loss": 3.9581,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12975845195409e-05,
+      "loss": 3.9599,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128919857203038e-05,
+      "loss": 3.9754,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128081262451986e-05,
+      "loss": 3.9562,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127242667700934e-05,
+      "loss": 3.9714,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126405710830255e-05,
+      "loss": 3.9675,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0246262550354,
+      "eval_runtime": 291.2514,
+      "eval_samples_per_second": 1310.178,
+      "eval_steps_per_second": 40.944,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125567116079203e-05,
+      "loss": 3.9564,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124728521328151e-05,
+      "loss": 3.9562,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123889926577099e-05,
+      "loss": 3.9574,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123051331826047e-05,
+      "loss": 3.9671,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122212737074995e-05,
+      "loss": 3.9675,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121374142323943e-05,
+      "loss": 3.9726,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1205371854532636e-05,
+      "loss": 3.959,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1196985907022116e-05,
+      "loss": 3.9619,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1188599959511596e-05,
+      "loss": 3.9642,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118021401200108e-05,
+      "loss": 3.9736,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117182806449056e-05,
+      "loss": 3.9564,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116344211698004e-05,
+      "loss": 3.9635,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115507254827325e-05,
+      "loss": 3.9693,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114668660076273e-05,
+      "loss": 3.9563,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113830065325221e-05,
+      "loss": 3.9565,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112991470574169e-05,
+      "loss": 3.9539,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1121528758231165e-05,
+      "loss": 3.9549,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1113142810720645e-05,
+      "loss": 3.9563,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1104756863210125e-05,
+      "loss": 3.9545,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109638729450334e-05,
+      "loss": 3.9549,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108800134699282e-05,
+      "loss": 3.9686,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.10796153994823e-05,
+      "loss": 3.9633,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107122945197178e-05,
+      "loss": 3.9559,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106284350446126e-05,
+      "loss": 3.9655,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105445755695074e-05,
+      "loss": 3.9719,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104607160944022e-05,
+      "loss": 3.9504,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.10376856619297e-05,
+      "loss": 3.9523,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102929971441918e-05,
+      "loss": 3.9646,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102093014571239e-05,
+      "loss": 3.9556,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.10125605770056e-05,
+      "loss": 3.9558,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100417462949508e-05,
+      "loss": 3.9547,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099578868198456e-05,
+      "loss": 3.9612,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098740273447404e-05,
+      "loss": 3.9667,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097901678696352e-05,
+      "loss": 3.9693,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0970647218256734e-05,
+      "loss": 3.9628,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0962261270746214e-05,
+      "loss": 3.9581,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0953875323235694e-05,
+      "loss": 3.9665,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0945489375725174e-05,
+      "loss": 3.9424,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0937103428214654e-05,
+      "loss": 3.9668,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0928717480704134e-05,
+      "loss": 3.948,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0920331533193614e-05,
+      "loss": 3.9516,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0911945585683094e-05,
+      "loss": 3.9593,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09035760169763e-05,
+      "loss": 3.9652,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089519006946578e-05,
+      "loss": 3.9585,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088680412195526e-05,
+      "loss": 3.959,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087841817444474e-05,
+      "loss": 3.9565,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087004860573796e-05,
+      "loss": 3.9527,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086167903703117e-05,
+      "loss": 3.9686,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085329308952065e-05,
+      "loss": 3.9384,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084492352081386e-05,
+      "loss": 3.9515,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083653757330334e-05,
+      "loss": 3.9666,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082815162579282e-05,
+      "loss": 3.962,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08197656782823e-05,
+      "loss": 3.9455,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081137973077178e-05,
+      "loss": 3.9533,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080299378326126e-05,
+      "loss": 3.9525,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079460783575074e-05,
+      "loss": 3.9457,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078622188824022e-05,
+      "loss": 3.9745,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07778359407297e-05,
+      "loss": 3.9372,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076944999321918e-05,
+      "loss": 3.9634,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0761064045708657e-05,
+      "loss": 3.9554,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0752678098198137e-05,
+      "loss": 3.9486,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0744292150687617e-05,
+      "loss": 3.9615,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073593896078456e-05,
+      "loss": 3.9455,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072755301327404e-05,
+      "loss": 3.9475,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071916706576352e-05,
+      "loss": 3.9509,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0710781118253e-05,
+      "loss": 3.9679,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0702395170742475e-05,
+      "loss": 3.9443,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0694009223231955e-05,
+      "loss": 3.9613,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06856560333289e-05,
+      "loss": 3.9394,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067727008581838e-05,
+      "loss": 3.95,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0668884138307866e-05,
+      "loss": 3.968,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0660498190797346e-05,
+      "loss": 3.9646,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0652112243286826e-05,
+      "loss": 3.9561,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0643726295776306e-05,
+      "loss": 3.9635,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063534034826578e-05,
+      "loss": 3.9707,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062695440075526e-05,
+      "loss": 3.9529,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061856845324474e-05,
+      "loss": 3.9651,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061018250573422e-05,
+      "loss": 3.9485,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060181293702743e-05,
+      "loss": 3.9457,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059342698951691e-05,
+      "loss": 3.9495,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058504104200639e-05,
+      "loss": 3.9555,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057665509449587e-05,
+      "loss": 3.9629,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0568285525789084e-05,
+      "loss": 3.9753,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0559899578278564e-05,
+      "loss": 3.9635,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0551513630768044e-05,
+      "loss": 3.947,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0543127683257524e-05,
+      "loss": 3.9509,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0534741735747004e-05,
+      "loss": 3.9624,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052637216704021e-05,
+      "loss": 3.9544,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051798621952969e-05,
+      "loss": 3.9521,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509600272019173e-05,
+      "loss": 3.9552,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501214324508653e-05,
+      "loss": 3.9476,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492828376998133e-05,
+      "loss": 3.9591,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484458808291345e-05,
+      "loss": 3.9473,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0476072860780825e-05,
+      "loss": 3.9618,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0467686913270305e-05,
+      "loss": 3.9583,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0459300965759785e-05,
+      "loss": 3.9532,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0450915018249265e-05,
+      "loss": 3.9479,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0442529070738745e-05,
+      "loss": 3.951,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0434143123228225e-05,
+      "loss": 3.9508,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0425773554521437e-05,
+      "loss": 3.9621,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0417387607010917e-05,
+      "loss": 3.9556,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0409001659500397e-05,
+      "loss": 3.9433,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0400615711989877e-05,
+      "loss": 3.9526,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0392246143283086e-05,
+      "loss": 3.9508,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383860195772566e-05,
+      "loss": 3.9559,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.037547424826205e-05,
+      "loss": 3.9498,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036708830075153e-05,
+      "loss": 3.9446,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035870235324101e-05,
+      "loss": 3.9574,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035031640573049e-05,
+      "loss": 3.9528,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0341930458219963e-05,
+      "loss": 3.9424,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0333544510709443e-05,
+      "loss": 3.9446,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325158563198926e-05,
+      "loss": 3.9455,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0316772615688406e-05,
+      "loss": 3.9606,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0308386668177886e-05,
+      "loss": 3.948,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0300000720667366e-05,
+      "loss": 3.9529,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291631151960575e-05,
+      "loss": 3.9595,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0283245204450055e-05,
+      "loss": 3.9686,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0274859256939535e-05,
+      "loss": 3.9541,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266489688232747e-05,
+      "loss": 3.941,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0258103740722227e-05,
+      "loss": 3.9566,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249717793211707e-05,
+      "loss": 3.955,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241331845701187e-05,
+      "loss": 3.9467,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0232945898190667e-05,
+      "loss": 3.948,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224559950680147e-05,
+      "loss": 3.9523,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216174003169627e-05,
+      "loss": 3.954,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.020780443446284e-05,
+      "loss": 3.9488,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019941848695232e-05,
+      "loss": 3.9576,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01910325394418e-05,
+      "loss": 3.9485,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.018266297073501e-05,
+      "loss": 3.9533,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.017427702322449e-05,
+      "loss": 3.9543,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.016589107571397e-05,
+      "loss": 3.9621,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157505128203452e-05,
+      "loss": 3.9543,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0149119180692932e-05,
+      "loss": 3.9425,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140733233182412e-05,
+      "loss": 3.9551,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132347285671892e-05,
+      "loss": 3.9593,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01239777169651e-05,
+      "loss": 3.9548,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.011559176945458e-05,
+      "loss": 3.9617,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010720582194406e-05,
+      "loss": 3.9482,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098819874433544e-05,
+      "loss": 3.9651,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0090433926923024e-05,
+      "loss": 3.9568,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0082047979412497e-05,
+      "loss": 3.9574,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073662031901977e-05,
+      "loss": 3.949,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065276084391457e-05,
+      "loss": 3.9622,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056890136880937e-05,
+      "loss": 3.9465,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004852056817415e-05,
+      "loss": 3.9458,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004013462066363e-05,
+      "loss": 3.9649,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003174867315311e-05,
+      "loss": 3.9493,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002336272564259e-05,
+      "loss": 3.9543,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.001497677813207e-05,
+      "loss": 3.9585,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.000659083062155e-05,
+      "loss": 3.9447,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.01918363571167,
+      "eval_runtime": 292.1231,
+      "eval_samples_per_second": 1306.268,
+      "eval_steps_per_second": 40.822,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.999820488311103e-05,
+      "loss": 3.9457,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989818935600513e-05,
+      "loss": 3.9463,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981432988089993e-05,
+      "loss": 3.9565,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973047040579473e-05,
+      "loss": 3.955,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964661093068953e-05,
+      "loss": 3.961,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956275145558432e-05,
+      "loss": 3.9511,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947889198047912e-05,
+      "loss": 3.951,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.993950325053739e-05,
+      "loss": 3.948,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.993111730302687e-05,
+      "loss": 3.9666,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.992273135551635e-05,
+      "loss": 3.9439,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.991434540800583e-05,
+      "loss": 3.9526,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.990595946049531e-05,
+      "loss": 3.9567,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989757351298479e-05,
+      "loss": 3.9499,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988918756547427e-05,
+      "loss": 3.9432,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880817996767478e-05,
+      "loss": 3.9413,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987243204925696e-05,
+      "loss": 3.9411,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.986404610174644e-05,
+      "loss": 3.9464,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.985566015423592e-05,
+      "loss": 3.9446,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98472742067254e-05,
+      "loss": 3.9407,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983890463801861e-05,
+      "loss": 3.9615,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983051869050809e-05,
+      "loss": 3.9513,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.982213274299757e-05,
+      "loss": 3.9459,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813763174290783e-05,
+      "loss": 3.9538,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805377226780262e-05,
+      "loss": 3.9606,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9796991279269742e-05,
+      "loss": 3.939,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788605331759222e-05,
+      "loss": 3.9443,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780219384248702e-05,
+      "loss": 3.9496,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771833436738182e-05,
+      "loss": 3.9512,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763447489227662e-05,
+      "loss": 3.9434,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755061541717146e-05,
+      "loss": 3.9438,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746675594206626e-05,
+      "loss": 3.9514,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97382896466961e-05,
+      "loss": 3.9558,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972990369918558e-05,
+      "loss": 3.958,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972151775167506e-05,
+      "loss": 3.9535,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.971313180416454e-05,
+      "loss": 3.9489,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704745856654022e-05,
+      "loss": 3.9511,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969637628794723e-05,
+      "loss": 3.9328,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968799034043671e-05,
+      "loss": 3.9554,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967960439292619e-05,
+      "loss": 3.9384,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967121844541567e-05,
+      "loss": 3.9414,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662848876708883e-05,
+      "loss": 3.9459,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654462929198363e-05,
+      "loss": 3.9584,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646093360491572e-05,
+      "loss": 3.9458,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637707412981052e-05,
+      "loss": 3.9441,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9629321465470532e-05,
+      "loss": 3.9502,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9620935517960012e-05,
+      "loss": 3.9392,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9612565949253225e-05,
+      "loss": 3.9563,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604180001742705e-05,
+      "loss": 3.9299,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9595794054232185e-05,
+      "loss": 3.9376,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9587408106721665e-05,
+      "loss": 3.9547,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9579022159211145e-05,
+      "loss": 3.9527,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570636211700625e-05,
+      "loss": 3.9373,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562250264190105e-05,
+      "loss": 3.9406,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9553864316679584e-05,
+      "loss": 3.9442,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545494747972797e-05,
+      "loss": 3.9305,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537108800462277e-05,
+      "loss": 3.9622,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528722852951757e-05,
+      "loss": 3.9335,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520336905441237e-05,
+      "loss": 3.9504,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9511967336734446e-05,
+      "loss": 3.9456,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.950358138922393e-05,
+      "loss": 3.936,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949519544171341e-05,
+      "loss": 3.9464,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.948680949420289e-05,
+      "loss": 3.9396,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947842354669237e-05,
+      "loss": 3.9391,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947003759918185e-05,
+      "loss": 3.9366,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.946165165167133e-05,
+      "loss": 3.9593,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945326570416081e-05,
+      "loss": 3.938,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.944489613545402e-05,
+      "loss": 3.9443,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94365101879435e-05,
+      "loss": 3.9317,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942812424043298e-05,
+      "loss": 3.9423,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941973829292246e-05,
+      "loss": 3.9557,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941136872421567e-05,
+      "loss": 3.9514,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940298277670515e-05,
+      "loss": 3.943,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394613207998363e-05,
+      "loss": 3.9559,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386227260487843e-05,
+      "loss": 3.9586,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377841312977323e-05,
+      "loss": 3.9411,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369455365466803e-05,
+      "loss": 3.952,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361069417956283e-05,
+      "loss": 3.9413,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935268347044576e-05,
+      "loss": 3.9361,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344313901738972e-05,
+      "loss": 3.9397,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9335927954228455e-05,
+      "loss": 3.9431,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327542006717935e-05,
+      "loss": 3.9511,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319156059207408e-05,
+      "loss": 3.9666,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310770111696888e-05,
+      "loss": 3.9542,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302400542990104e-05,
+      "loss": 3.9317,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294014595479584e-05,
+      "loss": 3.9454,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.928562864796906e-05,
+      "loss": 3.9493,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927724270045854e-05,
+      "loss": 3.9452,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.926885675294802e-05,
+      "loss": 3.9425,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.92604708054375e-05,
+      "loss": 3.9411,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.925208485792698e-05,
+      "loss": 3.9387,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243715289220193e-05,
+      "loss": 3.9465,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235329341709673e-05,
+      "loss": 3.9399,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9226943394199153e-05,
+      "loss": 3.9537,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218557446688633e-05,
+      "loss": 3.9428,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9210171499178113e-05,
+      "loss": 3.9486,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201801930471322e-05,
+      "loss": 3.9325,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193415982960802e-05,
+      "loss": 3.9419,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9185030035450285e-05,
+      "loss": 3.9437,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9176644087939765e-05,
+      "loss": 3.9494,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9168258140429245e-05,
+      "loss": 3.9502,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9159872192918725e-05,
+      "loss": 3.9323,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151502624211934e-05,
+      "loss": 3.9436,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143116676701414e-05,
+      "loss": 3.9385,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134747107994627e-05,
+      "loss": 3.9471,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126361160484106e-05,
+      "loss": 3.9368,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9117975212973586e-05,
+      "loss": 3.9349,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109589265463066e-05,
+      "loss": 3.9481,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101203317952546e-05,
+      "loss": 3.9378,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092817370442026e-05,
+      "loss": 3.9339,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084431422931506e-05,
+      "loss": 3.9344,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076045475420986e-05,
+      "loss": 3.9359,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.906765952791047e-05,
+      "loss": 3.9482,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905928995920368e-05,
+      "loss": 3.9385,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905090401169316e-05,
+      "loss": 3.9422,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904251806418264e-05,
+      "loss": 3.9502,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903413211667212e-05,
+      "loss": 3.9583,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025746169161595e-05,
+      "loss": 3.9411,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901737660045481e-05,
+      "loss": 3.9339,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900899065294429e-05,
+      "loss": 3.9418,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900060470543377e-05,
+      "loss": 3.9462,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992218757923244e-05,
+      "loss": 3.9356,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898384918921646e-05,
+      "loss": 3.9391,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897546324170594e-05,
+      "loss": 3.9391,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967077294195423e-05,
+      "loss": 3.9421,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958691346684896e-05,
+      "loss": 3.9418,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950305399174376e-05,
+      "loss": 3.9437,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8941919451663856e-05,
+      "loss": 3.9385,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.89335662617608e-05,
+      "loss": 3.9432,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925180314250285e-05,
+      "loss": 3.9397,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916794366739765e-05,
+      "loss": 3.9577,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908408419229245e-05,
+      "loss": 3.9441,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900022471718718e-05,
+      "loss": 3.9388,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891636524208198e-05,
+      "loss": 3.9367,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883250576697678e-05,
+      "loss": 3.9482,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887486462918716e-05,
+      "loss": 3.9455,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886649506048037e-05,
+      "loss": 3.9522,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885810911296985e-05,
+      "loss": 3.9386,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.884972316545933e-05,
+      "loss": 3.9539,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.884133721794881e-05,
+      "loss": 3.9446,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883295127043829e-05,
+      "loss": 3.9455,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882456532292777e-05,
+      "loss": 3.9413,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881617937541725e-05,
+      "loss": 3.9527,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807809806710462e-05,
+      "loss": 3.9414,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799423859199942e-05,
+      "loss": 3.9327,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791037911689422e-05,
+      "loss": 3.954,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782651964178902e-05,
+      "loss": 3.942,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774266016668382e-05,
+      "loss": 3.9423,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765896447961595e-05,
+      "loss": 3.9477,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757510500451075e-05,
+      "loss": 3.9354,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.014156341552734,
+      "eval_runtime": 290.875,
+      "eval_samples_per_second": 1311.873,
+      "eval_steps_per_second": 40.997,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8749124552940555e-05,
+      "loss": 3.9358,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8740738605430035e-05,
+      "loss": 3.9387,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8732352657919514e-05,
+      "loss": 3.9458,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8723966710408994e-05,
+      "loss": 3.9416,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715580762898474e-05,
+      "loss": 3.9544,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707194815387954e-05,
+      "loss": 3.9388,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698825246681167e-05,
+      "loss": 3.9435,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690455677974376e-05,
+      "loss": 3.9378,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8682069730463856e-05,
+      "loss": 3.9535,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8673683782953336e-05,
+      "loss": 3.9364,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8665297835442816e-05,
+      "loss": 3.9378,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8656911887932296e-05,
+      "loss": 3.9448,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.864852594042178e-05,
+      "loss": 3.9399,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640139992911252e-05,
+      "loss": 3.935,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631754045400732e-05,
+      "loss": 3.9368,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623384476693948e-05,
+      "loss": 3.9238,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8614998529183428e-05,
+      "loss": 3.9379,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8606612581672905e-05,
+      "loss": 3.9333,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8598226634162385e-05,
+      "loss": 3.9289,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8589840686651865e-05,
+      "loss": 3.9549,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8581454739141344e-05,
+      "loss": 3.9362,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8573085170434554e-05,
+      "loss": 3.943,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.856471560172777e-05,
+      "loss": 3.9395,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.855632965421725e-05,
+      "loss": 3.9519,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8547943706706726e-05,
+      "loss": 3.9283,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539557759196206e-05,
+      "loss": 3.9332,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531171811685686e-05,
+      "loss": 3.9408,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522785864175166e-05,
+      "loss": 3.9406,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514399916664646e-05,
+      "loss": 3.9306,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8506013969154126e-05,
+      "loss": 3.933,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.849762802164361e-05,
+      "loss": 3.9417,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489258452936818e-05,
+      "loss": 3.9441,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8480872505426298e-05,
+      "loss": 3.9469,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472486557915778e-05,
+      "loss": 3.9398,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464100610405258e-05,
+      "loss": 3.9359,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455714662894738e-05,
+      "loss": 3.9455,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447328715384218e-05,
+      "loss": 3.9214,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843895914667743e-05,
+      "loss": 3.9474,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843057319916691e-05,
+      "loss": 3.9265,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.842218725165639e-05,
+      "loss": 3.9352,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.841380130414587e-05,
+      "loss": 3.9348,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.840541535663535e-05,
+      "loss": 3.9468,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.839702940912483e-05,
+      "loss": 3.9394,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838864346161431e-05,
+      "loss": 3.9324,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380273892907523e-05,
+      "loss": 3.9396,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8371887945397003e-05,
+      "loss": 3.9334,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363501997886483e-05,
+      "loss": 3.9451,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8355116050375963e-05,
+      "loss": 3.9224,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.834674648166917e-05,
+      "loss": 3.9246,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338360534158655e-05,
+      "loss": 3.9441,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8329990965451864e-05,
+      "loss": 3.9433,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321605017941344e-05,
+      "loss": 3.9293,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313219070430824e-05,
+      "loss": 3.9243,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304833122920304e-05,
+      "loss": 3.9373,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296463554213516e-05,
+      "loss": 3.9184,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8288077606702996e-05,
+      "loss": 3.9462,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279691659192476e-05,
+      "loss": 3.9296,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271305711681956e-05,
+      "loss": 3.9395,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8262919764171436e-05,
+      "loss": 3.9347,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254533816660916e-05,
+      "loss": 3.9262,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8246147869150393e-05,
+      "loss": 3.9377,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8237761921639873e-05,
+      "loss": 3.9309,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229375974129353e-05,
+      "loss": 3.9253,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8220990026618833e-05,
+      "loss": 3.9281,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212604079108313e-05,
+      "loss": 3.9478,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204218131597793e-05,
+      "loss": 3.9322,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8195832184087273e-05,
+      "loss": 3.9308,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818746261538048e-05,
+      "loss": 3.9238,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8179076667869965e-05,
+      "loss": 3.9293,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8170690720359445e-05,
+      "loss": 3.9449,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8162304772848925e-05,
+      "loss": 3.9433,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8153935204142134e-05,
+      "loss": 3.9342,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145549256631614e-05,
+      "loss": 3.9437,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137163309121094e-05,
+      "loss": 3.9475,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128777361610574e-05,
+      "loss": 3.9362,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120407792903786e-05,
+      "loss": 3.9426,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8112021845393266e-05,
+      "loss": 3.9329,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103635897882746e-05,
+      "loss": 3.9252,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095249950372226e-05,
+      "loss": 3.9301,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086864002861706e-05,
+      "loss": 3.9336,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078478055351186e-05,
+      "loss": 3.9371,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070092107840666e-05,
+      "loss": 3.9569,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.806170616033015e-05,
+      "loss": 3.9433,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.805333659162336e-05,
+      "loss": 3.9217,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804495064411284e-05,
+      "loss": 3.935,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803656469660232e-05,
+      "loss": 3.9378,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028195127895527e-05,
+      "loss": 3.9376,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801980918038501e-05,
+      "loss": 3.9311,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801142323287449e-05,
+      "loss": 3.9319,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800303728536397e-05,
+      "loss": 3.9303,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.799465133785345e-05,
+      "loss": 3.9353,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.798628176914666e-05,
+      "loss": 3.9282,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797789582163614e-05,
+      "loss": 3.9417,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796950987412562e-05,
+      "loss": 3.9331,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961123926615103e-05,
+      "loss": 3.9387,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952737979104576e-05,
+      "loss": 3.9242,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7944352031594056e-05,
+      "loss": 3.9358,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7935966084083536e-05,
+      "loss": 3.9327,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927580136573016e-05,
+      "loss": 3.9382,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919194189062496e-05,
+      "loss": 3.9384,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.791082462035571e-05,
+      "loss": 3.9269,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.790243867284519e-05,
+      "loss": 3.9309,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.789405272533467e-05,
+      "loss": 3.9297,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.788566677782415e-05,
+      "loss": 3.9368,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.787728083031363e-05,
+      "loss": 3.9252,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786891126160684e-05,
+      "loss": 3.9292,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786054169290005e-05,
+      "loss": 3.9374,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.785215574538953e-05,
+      "loss": 3.9272,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.784376979787901e-05,
+      "loss": 3.9273,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.783538385036849e-05,
+      "loss": 3.9251,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.782699790285797e-05,
+      "loss": 3.922,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781861195534745e-05,
+      "loss": 3.9412,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810226007836933e-05,
+      "loss": 3.926,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801840060326413e-05,
+      "loss": 3.936,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7793454112815893e-05,
+      "loss": 3.9332,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785068165305373e-05,
+      "loss": 3.9483,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776682217794853e-05,
+      "loss": 3.9348,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768296270284333e-05,
+      "loss": 3.9233,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7759926701577542e-05,
+      "loss": 3.9319,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751557132870754e-05,
+      "loss": 3.9389,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743171185360234e-05,
+      "loss": 3.9229,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734801616653443e-05,
+      "loss": 3.9308,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726415669142923e-05,
+      "loss": 3.9268,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7718029721632403e-05,
+      "loss": 3.9403,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709643774121887e-05,
+      "loss": 3.9263,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701257826611367e-05,
+      "loss": 3.9354,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7692871879100847e-05,
+      "loss": 3.9284,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684485931590327e-05,
+      "loss": 3.9335,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676099984079807e-05,
+      "loss": 3.9258,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667730415373016e-05,
+      "loss": 3.9479,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659344467862496e-05,
+      "loss": 3.9377,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7650958520351976e-05,
+      "loss": 3.9263,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.764257257284146e-05,
+      "loss": 3.9296,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.763418662533094e-05,
+      "loss": 3.9349,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625817056624148e-05,
+      "loss": 3.9383,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617447487917357e-05,
+      "loss": 3.9443,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609061540406837e-05,
+      "loss": 3.9286,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760067559289632e-05,
+      "loss": 3.9418,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.75922896453858e-05,
+      "loss": 3.9334,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758390369787528e-05,
+      "loss": 3.9378,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757551775036476e-05,
+      "loss": 3.9309,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7567131802854233e-05,
+      "loss": 3.9396,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558745855343713e-05,
+      "loss": 3.9318,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7550359907833197e-05,
+      "loss": 3.9274,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7541990339126406e-05,
+      "loss": 3.9417,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7533604391615886e-05,
+      "loss": 3.934,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525218444105366e-05,
+      "loss": 3.9348,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7516832496594846e-05,
+      "loss": 3.9368,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508446549084326e-05,
+      "loss": 3.9278,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.00921106338501,
+      "eval_runtime": 289.5674,
+      "eval_samples_per_second": 1317.797,
+      "eval_steps_per_second": 41.182,
+      "step": 1373760
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 5.636333087176712e+17,
+  "trial_name": null,
+  "trial_params": null
+}