diff --git "a/checkpoint-2365920/trainer_state.json" "b/checkpoint-2365920/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2365920/trainer_state.json"
@@ -0,0 +1,27993 @@
+{
+  "best_metric": 3.980487108230591,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl2/lstm/0/checkpoints/checkpoint-2365920",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2365920,
+  "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": 10.8196,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5546,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0588,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9975,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.948,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9267,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7497,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6233,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5276,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.449,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.4031,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.344,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2792,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2144,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1629,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.1059,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0645,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0259,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.98,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9363,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.9066,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8626,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8368,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.799,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7804,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7417,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7184,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6894,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6715,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6447,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6314,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.6139,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5886,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5732,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5575,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5351,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5314,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4978,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4833,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4548,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4532,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4371,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4339,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4069,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3969,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3948,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3864,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3737,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3556,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3407,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.357,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3441,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3228,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.3016,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.3115,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.2911,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530501591037e-05,
+      "loss": 5.2943,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952211564352648e-05,
+      "loss": 5.265,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2777,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2603,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2428,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2297,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2418,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.199,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.2142,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.2204,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.204,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.195,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1705,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1663,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1631,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.174,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1529,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1619,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1523,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.1499,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936284815604152e-05,
+      "loss": 5.1269,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354462208531e-05,
+      "loss": 5.1248,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934607626102048e-05,
+      "loss": 5.1154,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1125,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1085,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0933,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0954,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0925,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.068,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0763,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0739,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0752,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.055,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0527,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0386,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0347,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0358,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0332,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0241,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.027,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0176,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9996,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9973,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9989,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9954,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9989,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9804,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9724,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9768,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9575,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9627,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9508,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9528,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.94,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9416,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9314,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9353,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9299,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9197,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9306,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9266,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9163,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9031,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9036,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9011,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.9034,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8975,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8861,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8913,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8778,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8808,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8659,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8734,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.862,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.8718,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8703,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8673,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8442,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8504,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8477,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8494,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8377,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8434,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.832,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8467,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8147,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8336,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8186,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.814,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8138,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.798,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8026,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8112,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7969,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.763216018676758,
+      "eval_runtime": 291.1721,
+      "eval_samples_per_second": 1310.534,
+      "eval_steps_per_second": 40.955,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7958,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7904,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.8009,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7898,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7963,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7727,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7797,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7651,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7687,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7717,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7684,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7728,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7506,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7642,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7473,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7498,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.753,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.743,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.7451,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.7585,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.735,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7409,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7343,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7436,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7331,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7313,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7142,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7229,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.7097,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.7204,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.7163,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.714,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.7175,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7071,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.714,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7061,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6957,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6884,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.6798,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.6873,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6871,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6923,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6764,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6781,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.6816,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.6744,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.68,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.665,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6578,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6848,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6778,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6643,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6594,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6667,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6511,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.6678,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.6387,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.6644,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.6559,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.6392,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.6358,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6519,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6143,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6365,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6459,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.639,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.6287,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.616,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6203,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6238,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6292,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.6198,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6269,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.6313,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122235667400224e-05,
+      "loss": 4.6275,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6112,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.6157,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809711058247612e-05,
+      "loss": 4.6156,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.6082,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808033868745508e-05,
+      "loss": 4.6117,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807195273994456e-05,
+      "loss": 4.6031,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806356679243404e-05,
+      "loss": 4.6102,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805518084492352e-05,
+      "loss": 4.6101,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.5873,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.602,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.5998,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6075,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.5976,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.5925,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.5881,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5815,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5876,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.5908,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.5876,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.5924,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.5844,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.5766,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929440768676916e-05,
+      "loss": 4.5735,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921054821166396e-05,
+      "loss": 4.5786,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.583,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904282926145356e-05,
+      "loss": 4.5831,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895913357438565e-05,
+      "loss": 4.575,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887527409928045e-05,
+      "loss": 4.5657,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879157841221254e-05,
+      "loss": 4.5766,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870771893710734e-05,
+      "loss": 4.5581,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862385946200214e-05,
+      "loss": 4.5674,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853999998689694e-05,
+      "loss": 4.5611,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784561405117918e-05,
+      "loss": 4.5627,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783722810366866e-05,
+      "loss": 4.551,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782884215615814e-05,
+      "loss": 4.5613,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.55,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.5542,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5574,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.547,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5678,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777854284989875e-05,
+      "loss": 4.5631,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.5561,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5446,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5418,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5493,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.5441,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5552,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.5352,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5428,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.5415,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.5461,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.5264,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5391,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5351,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5364,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5417,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5448,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5223,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.531,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.5337,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5293,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760250346739275e-05,
+      "loss": 4.5264,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759411751988223e-05,
+      "loss": 4.5305,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.5235,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.5396,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.5139,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5303,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.517,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5158,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.5161,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5093,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518676749895016e-05,
+      "loss": 4.5092,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510290802384496e-05,
+      "loss": 4.5149,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501904854873976e-05,
+      "loss": 4.5095,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.47607421875,
+      "eval_runtime": 291.0564,
+      "eval_samples_per_second": 1311.055,
+      "eval_steps_per_second": 40.971,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493518907363456e-05,
+      "loss": 4.5083,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485132959852936e-05,
+      "loss": 4.5002,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476747012342416e-05,
+      "loss": 4.5195,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468361064831896e-05,
+      "loss": 4.5069,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459975117321376e-05,
+      "loss": 4.5188,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745158916981085e-05,
+      "loss": 4.4955,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744320322230033e-05,
+      "loss": 4.5033,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743481727478981e-05,
+      "loss": 4.4903,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742643132727929e-05,
+      "loss": 4.5007,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741804537976877e-05,
+      "loss": 4.4998,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740965943225825e-05,
+      "loss": 4.496,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401273484747736e-05,
+      "loss": 4.5094,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392903916040945e-05,
+      "loss": 4.4829,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384517968530425e-05,
+      "loss": 4.5007,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376132021019905e-05,
+      "loss": 4.4848,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367746073509385e-05,
+      "loss": 4.4844,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359360125998865e-05,
+      "loss": 4.4859,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350974178488345e-05,
+      "loss": 4.4927,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342588230977825e-05,
+      "loss": 4.4833,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334218662271034e-05,
+      "loss": 4.5025,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325832714760514e-05,
+      "loss": 4.485,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731746314605372e-05,
+      "loss": 4.4853,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73090771985432e-05,
+      "loss": 4.4828,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730069125103269e-05,
+      "loss": 4.496,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729230530352217e-05,
+      "loss": 4.4816,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728391935601165e-05,
+      "loss": 4.4846,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727553340850113e-05,
+      "loss": 4.4714,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726714746099061e-05,
+      "loss": 4.4759,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725876151348009e-05,
+      "loss": 4.4652,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725037556596957e-05,
+      "loss": 4.4799,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724198961845904e-05,
+      "loss": 4.4806,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723360367094852e-05,
+      "loss": 4.4751,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225217723438e-05,
+      "loss": 4.4836,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721684815473121e-05,
+      "loss": 4.4715,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4727,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.4748,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4661,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718332074349287e-05,
+      "loss": 4.457,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717493479598235e-05,
+      "loss": 4.4499,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716654884847183e-05,
+      "loss": 4.4585,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715816290096131e-05,
+      "loss": 4.4618,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714977695345079e-05,
+      "loss": 4.4659,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141407384743996e-05,
+      "loss": 4.4562,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133021437233476e-05,
+      "loss": 4.451,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.4636,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116265921016165e-05,
+      "loss": 4.4549,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107879973505645e-05,
+      "loss": 4.4556,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099494025995125e-05,
+      "loss": 4.4496,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091108078484605e-05,
+      "loss": 4.4382,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708272213097409e-05,
+      "loss": 4.4686,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.4619,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4478,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4475,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704919471973574e-05,
+      "loss": 4.4566,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704080877222522e-05,
+      "loss": 4.4364,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4571,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702403687720418e-05,
+      "loss": 4.4307,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701565092969366e-05,
+      "loss": 4.4516,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700726498218314e-05,
+      "loss": 4.4535,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699887903467262e-05,
+      "loss": 4.4362,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699052584476956e-05,
+      "loss": 4.4289,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4472,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4166,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965368002238005e-05,
+      "loss": 4.4354,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956982054727485e-05,
+      "loss": 4.4405,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948596107216965e-05,
+      "loss": 4.441,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940210159706445e-05,
+      "loss": 4.4337,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931824212195925e-05,
+      "loss": 4.4187,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69234382646854e-05,
+      "loss": 4.4227,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691505231717488e-05,
+      "loss": 4.424,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906682748468094e-05,
+      "loss": 4.4353,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68983131797613e-05,
+      "loss": 4.426,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688992723225078e-05,
+      "loss": 4.4357,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688154128474026e-05,
+      "loss": 4.4393,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687315533722974e-05,
+      "loss": 4.4362,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686476938971923e-05,
+      "loss": 4.4179,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68563834422087e-05,
+      "loss": 4.4248,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684799749469818e-05,
+      "loss": 4.431,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683961154718766e-05,
+      "loss": 4.4182,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683124197848087e-05,
+      "loss": 4.4238,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682285603097035e-05,
+      "loss": 4.4194,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681447008345983e-05,
+      "loss": 4.4189,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680608413594931e-05,
+      "loss": 4.4261,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679769818843879e-05,
+      "loss": 4.4045,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789328619732e-05,
+      "loss": 4.4217,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678094267222148e-05,
+      "loss": 4.4136,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677255672471097e-05,
+      "loss": 4.4259,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676417077720045e-05,
+      "loss": 4.4151,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675578482968993e-05,
+      "loss": 4.4142,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747415260983136e-05,
+      "loss": 4.4084,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739029313472616e-05,
+      "loss": 4.4044,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730643365962096e-05,
+      "loss": 4.405,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722257418451576e-05,
+      "loss": 4.4123,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713887849744785e-05,
+      "loss": 4.4207,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705501902234265e-05,
+      "loss": 4.4151,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697115954723745e-05,
+      "loss": 4.4073,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688730007213225e-05,
+      "loss": 4.4021,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680360438506434e-05,
+      "loss": 4.4025,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667197449099592e-05,
+      "loss": 4.4042,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635885434854e-05,
+      "loss": 4.4135,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665520259597488e-05,
+      "loss": 4.4086,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664683302726809e-05,
+      "loss": 4.4038,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3942,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630093889854515e-05,
+      "loss": 4.4103,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621707942343995e-05,
+      "loss": 4.3909,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.4001,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.392,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3979,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3878,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.399,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657139225728087e-05,
+      "loss": 4.3888,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563006309770355e-05,
+      "loss": 4.3855,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554620362259835e-05,
+      "loss": 4.3996,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546234414749315e-05,
+      "loss": 4.3827,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537848467238795e-05,
+      "loss": 4.4053,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529478898532004e-05,
+      "loss": 4.4016,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521092951021484e-05,
+      "loss": 4.3957,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512707003510964e-05,
+      "loss": 4.3858,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3856,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3867,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3863,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647917959227261e-05,
+      "loss": 4.396,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647079364476209e-05,
+      "loss": 4.3784,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646240769725157e-05,
+      "loss": 4.3884,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645402174974105e-05,
+      "loss": 4.3878,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644563580223053e-05,
+      "loss": 4.3889,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643724985472001e-05,
+      "loss": 4.376,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642886390720949e-05,
+      "loss": 4.3838,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642047795969897e-05,
+      "loss": 4.3818,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641209201218845e-05,
+      "loss": 4.3848,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640370606467793e-05,
+      "loss": 4.3871,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639533649597114e-05,
+      "loss": 4.3935,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638695054846062e-05,
+      "loss": 4.3692,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63785646009501e-05,
+      "loss": 4.3825,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637017865343958e-05,
+      "loss": 4.3858,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636179270592906e-05,
+      "loss": 4.3795,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635340675841854e-05,
+      "loss": 4.3769,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3858,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633665124220124e-05,
+      "loss": 4.3753,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632826529469072e-05,
+      "loss": 4.3918,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63198793471802e-05,
+      "loss": 4.3664,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311509778473406e-05,
+      "loss": 4.3813,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303123830962886e-05,
+      "loss": 4.3743,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294737883452366e-05,
+      "loss": 4.3722,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286351935941846e-05,
+      "loss": 4.3696,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277965988431326e-05,
+      "loss": 4.3664,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269580040920806e-05,
+      "loss": 4.3625,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261194093410286e-05,
+      "loss": 4.3717,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252824524703495e-05,
+      "loss": 4.3676,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.3395867347717285,
+      "eval_runtime": 287.9323,
+      "eval_samples_per_second": 1325.28,
+      "eval_steps_per_second": 41.416,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244438577192975e-05,
+      "loss": 4.3731,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623605262968246e-05,
+      "loss": 4.3578,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622766668217194e-05,
+      "loss": 4.3766,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621928073466142e-05,
+      "loss": 4.3644,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210894787150895e-05,
+      "loss": 4.3811,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202508839640374e-05,
+      "loss": 4.3579,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3648,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3501,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177367375912543e-05,
+      "loss": 4.3651,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168981428402023e-05,
+      "loss": 4.3624,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.3591,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.3656,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61438399646742e-05,
+      "loss": 4.3518,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613545401716368e-05,
+      "loss": 4.3652,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3516,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3458,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.3522,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.3555,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.3538,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.3687,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607676876339377e-05,
+      "loss": 4.3532,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606838281588325e-05,
+      "loss": 4.3591,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605999686837273e-05,
+      "loss": 4.3497,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605162729966594e-05,
+      "loss": 4.3618,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604324135215542e-05,
+      "loss": 4.3503,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60348554046449e-05,
+      "loss": 4.3547,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602646945713438e-05,
+      "loss": 4.3439,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601809988842759e-05,
+      "loss": 4.3407,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600971394091707e-05,
+      "loss": 4.3418,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600132799340655e-05,
+      "loss": 4.3483,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599294204589603e-05,
+      "loss": 4.3513,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598455609838551e-05,
+      "loss": 4.3472,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597617015087499e-05,
+      "loss": 4.3571,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596778420336447e-05,
+      "loss": 4.3467,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3423,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595102868714716e-05,
+      "loss": 4.3465,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3467,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3292,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3242,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3322,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3421,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590072938088778e-05,
+      "loss": 4.3415,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892359812180986e-05,
+      "loss": 4.3364,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883973864670466e-05,
+      "loss": 4.3265,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875587917159946e-05,
+      "loss": 4.3418,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867201969649426e-05,
+      "loss": 4.334,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858816022138906e-05,
+      "loss": 4.3338,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585043007462838e-05,
+      "loss": 4.3268,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3124,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3546,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825288610900555e-05,
+      "loss": 4.3382,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816902663390035e-05,
+      "loss": 4.331,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3307,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3332,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3215,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3356,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774989304641164e-05,
+      "loss": 4.3163,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766603357130644e-05,
+      "loss": 4.3291,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758217409620124e-05,
+      "loss": 4.3402,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3201,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3098,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573307594589229e-05,
+      "loss": 4.3363,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572468999838177e-05,
+      "loss": 4.3019,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571632042967499e-05,
+      "loss": 4.3153,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570793448216447e-05,
+      "loss": 4.3281,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569954853465395e-05,
+      "loss": 4.327,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569116258714343e-05,
+      "loss": 4.3188,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568277663963291e-05,
+      "loss": 4.3115,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567439069212239e-05,
+      "loss": 4.3031,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566600474461187e-05,
+      "loss": 4.3094,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565761879710135e-05,
+      "loss": 4.3256,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564923284959083e-05,
+      "loss": 4.3101,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3255,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3218,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3296,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561570543835248e-05,
+      "loss": 4.3104,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560731949084196e-05,
+      "loss": 4.3113,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559893354333144e-05,
+      "loss": 4.3214,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559054759582092e-05,
+      "loss": 4.3053,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558219440591786e-05,
+      "loss": 4.3199,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557380845840734e-05,
+      "loss": 4.3082,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556542251089682e-05,
+      "loss": 4.3106,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55570365633863e-05,
+      "loss": 4.3177,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554866699467951e-05,
+      "loss": 4.2995,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554028104716899e-05,
+      "loss": 4.3114,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553189509965847e-05,
+      "loss": 4.3096,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552350915214795e-05,
+      "loss": 4.3163,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551513958344116e-05,
+      "loss": 4.3042,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550675363593064e-05,
+      "loss": 4.3085,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549836768842013e-05,
+      "loss": 4.3005,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548998174090961e-05,
+      "loss": 4.2977,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548159579339909e-05,
+      "loss": 4.3007,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547320984588857e-05,
+      "loss": 4.3054,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546482389837804e-05,
+      "loss": 4.3161,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545643795086752e-05,
+      "loss": 4.3137,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448052003357e-05,
+      "loss": 4.3038,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439682434650216e-05,
+      "loss": 4.2948,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543129648713969e-05,
+      "loss": 4.2976,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422926918432905e-05,
+      "loss": 4.3031,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414540970922385e-05,
+      "loss": 4.3075,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406155023411865e-05,
+      "loss": 4.3073,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397769075901345e-05,
+      "loss": 4.3011,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538939950719456e-05,
+      "loss": 4.2909,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538101355968404e-05,
+      "loss": 4.3112,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372627612173514e-05,
+      "loss": 4.284,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364241664662994e-05,
+      "loss": 4.3026,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355855717152473e-05,
+      "loss": 4.2894,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347469769641953e-05,
+      "loss": 4.2971,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533908382213143e-05,
+      "loss": 4.2884,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533069787462091e-05,
+      "loss": 4.2963,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532232830591412e-05,
+      "loss": 4.2909,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139423584036e-05,
+      "loss": 4.2848,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530555641089308e-05,
+      "loss": 4.2982,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52971868421863e-05,
+      "loss": 4.2853,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.3044,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528041494716526e-05,
+      "loss": 4.3057,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527202899965474e-05,
+      "loss": 4.3009,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526364305214422e-05,
+      "loss": 4.2879,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52552571046337e-05,
+      "loss": 4.2877,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524687115712318e-05,
+      "loss": 4.2863,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523848520961266e-05,
+      "loss": 4.289,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523011564090587e-05,
+      "loss": 4.3001,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522172969339535e-05,
+      "loss": 4.276,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521334374588483e-05,
+      "loss": 4.2916,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520495779837431e-05,
+      "loss": 4.2884,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519657185086379e-05,
+      "loss": 4.2955,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518818590335327e-05,
+      "loss": 4.2812,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517981633464648e-05,
+      "loss": 4.2903,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517143038713596e-05,
+      "loss": 4.2814,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516304443962544e-05,
+      "loss": 4.2952,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515465849211492e-05,
+      "loss": 4.2893,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51462725446044e-05,
+      "loss": 4.2983,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513790297589761e-05,
+      "loss": 4.2725,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512951702838709e-05,
+      "loss": 4.2911,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512113108087657e-05,
+      "loss": 4.2917,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112745133366045e-05,
+      "loss": 4.28,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104359185855525e-05,
+      "loss": 4.2839,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095973238345005e-05,
+      "loss": 4.294,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508760366963822e-05,
+      "loss": 4.2843,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50792177221277e-05,
+      "loss": 4.3003,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507083177461718e-05,
+      "loss": 4.2721,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506244582710666e-05,
+      "loss": 4.2902,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054076258399876e-05,
+      "loss": 4.2772,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504569031088935e-05,
+      "loss": 4.2813,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503730436337883e-05,
+      "loss": 4.2804,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502891841586831e-05,
+      "loss": 4.2735,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502053246835779e-05,
+      "loss": 4.2709,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012162899651e-05,
+      "loss": 4.2783,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500377695214048e-05,
+      "loss": 4.2783,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.257771968841553,
+      "eval_runtime": 288.3054,
+      "eval_samples_per_second": 1323.565,
+      "eval_steps_per_second": 41.362,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499539100462996e-05,
+      "loss": 4.2781,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498700505711944e-05,
+      "loss": 4.2686,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978619109608925e-05,
+      "loss": 4.2879,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970233162098405e-05,
+      "loss": 4.275,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961847214587885e-05,
+      "loss": 4.287,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953461267077365e-05,
+      "loss": 4.2754,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945075319566845e-05,
+      "loss": 4.2746,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936689372056325e-05,
+      "loss": 4.2628,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928303424545805e-05,
+      "loss": 4.2729,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919917477035285e-05,
+      "loss": 4.2745,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911531529524765e-05,
+      "loss": 4.2729,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490314558201424e-05,
+      "loss": 4.2748,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894776013307454e-05,
+      "loss": 4.2643,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886390065796934e-05,
+      "loss": 4.2786,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487800411828641e-05,
+      "loss": 4.2631,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486961817077589e-05,
+      "loss": 4.2602,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861232223265373e-05,
+      "loss": 4.2649,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852846275754853e-05,
+      "loss": 4.2712,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484446032824433e-05,
+      "loss": 4.2661,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483607438073381e-05,
+      "loss": 4.2811,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482768843322329e-05,
+      "loss": 4.2697,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481930248571277e-05,
+      "loss": 4.2773,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481091653820225e-05,
+      "loss": 4.2629,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480253059069173e-05,
+      "loss": 4.2752,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479417740078867e-05,
+      "loss": 4.2646,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478579145327815e-05,
+      "loss": 4.2722,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477740550576763e-05,
+      "loss": 4.2587,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476901955825711e-05,
+      "loss": 4.2589,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476063361074659e-05,
+      "loss": 4.2558,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475224766323607e-05,
+      "loss": 4.264,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474386171572556e-05,
+      "loss": 4.2681,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473549214701877e-05,
+      "loss": 4.2659,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472710619950825e-05,
+      "loss": 4.2708,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471872025199773e-05,
+      "loss": 4.2598,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710350683290936e-05,
+      "loss": 4.2617,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701964735780416e-05,
+      "loss": 4.2692,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693578788269896e-05,
+      "loss": 4.266,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685192840759376e-05,
+      "loss": 4.2444,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676806893248856e-05,
+      "loss": 4.2508,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668420945738336e-05,
+      "loss": 4.2482,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660034998227816e-05,
+      "loss": 4.2539,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2655,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643263103206776e-05,
+      "loss": 4.253,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634877155696256e-05,
+      "loss": 4.2482,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626491208185736e-05,
+      "loss": 4.2588,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618105260675215e-05,
+      "loss": 4.2527,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609735691968425e-05,
+      "loss": 4.2565,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460136612326164e-05,
+      "loss": 4.2495,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459298017575112e-05,
+      "loss": 4.2274,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584594228240594e-05,
+      "loss": 4.2755,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576208280730074e-05,
+      "loss": 4.259,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567822333219554e-05,
+      "loss": 4.2565,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559436385709033e-05,
+      "loss": 4.2504,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551050438198513e-05,
+      "loss": 4.2506,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454266449068799e-05,
+      "loss": 4.2442,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534311300784945e-05,
+      "loss": 4.2532,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525925353274425e-05,
+      "loss": 4.2419,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45175394057639e-05,
+      "loss": 4.2484,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450915345825338e-05,
+      "loss": 4.2633,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450076751074286e-05,
+      "loss": 4.2411,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449238156323234e-05,
+      "loss": 4.2339,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448399561572182e-05,
+      "loss": 4.2588,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447562604701503e-05,
+      "loss": 4.2246,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446724009950451e-05,
+      "loss": 4.2405,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445885415199399e-05,
+      "loss": 4.2476,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445046820448347e-05,
+      "loss": 4.249,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444208225697295e-05,
+      "loss": 4.2411,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443369630946243e-05,
+      "loss": 4.2376,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2222,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416924414441394e-05,
+      "loss": 4.2347,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44085548457346e-05,
+      "loss": 4.2458,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440018527702781e-05,
+      "loss": 4.2365,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439179932951729e-05,
+      "loss": 4.2441,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438341338200677e-05,
+      "loss": 4.2489,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437502743449625e-05,
+      "loss": 4.2569,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2357,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2318,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2482,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2277,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2483,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.2352,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316358559530076e-05,
+      "loss": 4.2351,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307972612019556e-05,
+      "loss": 4.2453,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2253,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291200716998516e-05,
+      "loss": 4.2354,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2324,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274445200781205e-05,
+      "loss": 4.2432,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266059253270685e-05,
+      "loss": 4.2286,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257673305760165e-05,
+      "loss": 4.2381,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249287358249645e-05,
+      "loss": 4.2259,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240901410739125e-05,
+      "loss": 4.2282,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232531842032334e-05,
+      "loss": 4.2237,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422414589452182e-05,
+      "loss": 4.235,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42157599470113e-05,
+      "loss": 4.2429,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420737399950078e-05,
+      "loss": 4.2377,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419898805199026e-05,
+      "loss": 4.2376,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419061848328347e-05,
+      "loss": 4.2235,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418223253577295e-05,
+      "loss": 4.2263,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417384658826243e-05,
+      "loss": 4.2294,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2376,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2338,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414868874573086e-05,
+      "loss": 4.2281,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414031917702408e-05,
+      "loss": 4.2223,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2396,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.2134,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411516133449252e-05,
+      "loss": 4.2302,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106775386982e-05,
+      "loss": 4.2196,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409840581827521e-05,
+      "loss": 4.2271,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.2181,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2267,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2185,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.2185,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2264,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.2183,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40397205645053e-05,
+      "loss": 4.2344,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031334616994777e-05,
+      "loss": 4.2361,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022948669484257e-05,
+      "loss": 4.2284,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401456272197374e-05,
+      "loss": 4.2246,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400617677446322e-05,
+      "loss": 4.2153,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39977908269527e-05,
+      "loss": 4.2207,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398940487944218e-05,
+      "loss": 4.22,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.236,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397264936322487e-05,
+      "loss": 4.2052,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396426341571435e-05,
+      "loss": 4.223,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395587746820383e-05,
+      "loss": 4.221,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394750789949704e-05,
+      "loss": 4.2262,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393912195198652e-05,
+      "loss": 4.2172,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930736004476e-05,
+      "loss": 4.2202,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.2141,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39139804882587e-05,
+      "loss": 4.227,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.2244,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389720859323766e-05,
+      "loss": 4.2265,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388882264572714e-05,
+      "loss": 4.209,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2242,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2257,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2149,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2165,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846925665781995e-05,
+      "loss": 4.2263,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838539718271475e-05,
+      "loss": 4.2158,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2387,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.2088,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.2216,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2091,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796609980718875e-05,
+      "loss": 4.2166,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788224033208355e-05,
+      "loss": 4.2178,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2088,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2093,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.2046,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2173,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.20110559463501,
+      "eval_runtime": 297.2997,
+      "eval_samples_per_second": 1283.523,
+      "eval_steps_per_second": 40.111,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.2185,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.2042,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2204,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2164,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371279964202487e-05,
+      "loss": 4.2198,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370441369451435e-05,
+      "loss": 4.2088,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369602774700383e-05,
+      "loss": 4.2112,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368764179949331e-05,
+      "loss": 4.1984,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367925585198279e-05,
+      "loss": 4.2089,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367086990447227e-05,
+      "loss": 4.2107,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366248395696175e-05,
+      "loss": 4.2139,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365409800945123e-05,
+      "loss": 4.2071,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364572844074444e-05,
+      "loss": 4.2075,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363734249323392e-05,
+      "loss": 4.2091,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36289565457234e-05,
+      "loss": 4.2022,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362057059821288e-05,
+      "loss": 4.1942,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612201029506086e-05,
+      "loss": 4.2019,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603815081995566e-05,
+      "loss": 4.2083,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359542913448505e-05,
+      "loss": 4.2044,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358704318697453e-05,
+      "loss": 4.2181,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357865723946401e-05,
+      "loss": 4.2095,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357028767075722e-05,
+      "loss": 4.2135,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35619017232467e-05,
+      "loss": 4.2011,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355351577573618e-05,
+      "loss": 4.2113,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354514620702939e-05,
+      "loss": 4.2036,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353676025951887e-05,
+      "loss": 4.2133,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352837431200835e-05,
+      "loss": 4.1984,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351998836449783e-05,
+      "loss": 4.1972,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351160241698731e-05,
+      "loss": 4.1952,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350321646947679e-05,
+      "loss": 4.2047,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349483052196627e-05,
+      "loss": 4.2031,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348644457445575e-05,
+      "loss": 4.2063,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347805862694523e-05,
+      "loss": 4.2079,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346967267943471e-05,
+      "loss": 4.2036,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346128673192419e-05,
+      "loss": 4.2002,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345290078441367e-05,
+      "loss": 4.2079,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344453121570688e-05,
+      "loss": 4.2038,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343614526819636e-05,
+      "loss": 4.1882,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342775932068584e-05,
+      "loss": 4.1897,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341937337317532e-05,
+      "loss": 4.187,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341100380446853e-05,
+      "loss": 4.1891,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340261785695801e-05,
+      "loss": 4.2056,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339423190944749e-05,
+      "loss": 4.1959,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385845961936975e-05,
+      "loss": 4.1897,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377460014426455e-05,
+      "loss": 4.2005,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33691068245234e-05,
+      "loss": 4.1912,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336072087701287e-05,
+      "loss": 4.1988,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335233492950235e-05,
+      "loss": 4.1895,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334394898199183e-05,
+      "loss": 4.1677,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333556303448131e-05,
+      "loss": 4.2178,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332717708697079e-05,
+      "loss": 4.1963,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331879113946027e-05,
+      "loss": 4.2019,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331040519194975e-05,
+      "loss": 4.1877,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330203562324296e-05,
+      "loss": 4.198,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329366605453618e-05,
+      "loss": 4.1852,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328528010702566e-05,
+      "loss": 4.1906,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327689415951514e-05,
+      "loss": 4.1903,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326850821200462e-05,
+      "loss": 4.1892,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32601222644941e-05,
+      "loss": 4.208,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325173631698358e-05,
+      "loss": 4.1828,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324335036947306e-05,
+      "loss": 4.1744,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323496442196254e-05,
+      "loss": 4.2015,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322657847445202e-05,
+      "loss": 4.1681,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321820890574523e-05,
+      "loss": 4.1775,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209822958234707e-05,
+      "loss": 4.1937,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201437010724187e-05,
+      "loss": 4.194,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193051063213666e-05,
+      "loss": 4.1812,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184681494506876e-05,
+      "loss": 4.1861,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317629554699636e-05,
+      "loss": 4.165,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31679423570933e-05,
+      "loss": 4.1814,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315955640958278e-05,
+      "loss": 4.1894,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315117046207226e-05,
+      "loss": 4.1783,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314278451456174e-05,
+      "loss": 4.1884,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313439856705122e-05,
+      "loss": 4.1901,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31260126195407e-05,
+      "loss": 4.2021,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311762667203018e-05,
+      "loss": 4.1771,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310924072451966e-05,
+      "loss": 4.1822,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310085477700914e-05,
+      "loss": 4.1924,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309246882949862e-05,
+      "loss": 4.1733,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30840828819881e-05,
+      "loss": 4.1881,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307569693447758e-05,
+      "loss": 4.1829,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306731098696706e-05,
+      "loss": 4.1807,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058941418260276e-05,
+      "loss": 4.1922,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050555470749756e-05,
+      "loss": 4.1715,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042169523239236e-05,
+      "loss": 4.1758,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303378357572871e-05,
+      "loss": 4.1825,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302539762821819e-05,
+      "loss": 4.1885,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301701168070767e-05,
+      "loss": 4.1742,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008642112000885e-05,
+      "loss": 4.183,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300025616449036e-05,
+      "loss": 4.1662,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299187021697984e-05,
+      "loss": 4.1761,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298348426946932e-05,
+      "loss": 4.1716,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29750983219588e-05,
+      "loss": 4.179,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966712374448285e-05,
+      "loss": 4.1921,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958326426937765e-05,
+      "loss": 4.1804,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949956858230974e-05,
+      "loss": 4.1811,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941570910720454e-05,
+      "loss": 4.1684,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933184963209934e-05,
+      "loss": 4.1713,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924799015699413e-05,
+      "loss": 4.1744,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291642944699262e-05,
+      "loss": 4.1852,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29080434994821e-05,
+      "loss": 4.1773,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289967393077531e-05,
+      "loss": 4.1816,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289128798326479e-05,
+      "loss": 4.1698,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288290203575427e-05,
+      "loss": 4.1844,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287451608824375e-05,
+      "loss": 4.1628,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286614651953697e-05,
+      "loss": 4.1776,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285776057202645e-05,
+      "loss": 4.162,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284937462451593e-05,
+      "loss": 4.1793,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284098867700541e-05,
+      "loss": 4.1664,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283260272949489e-05,
+      "loss": 4.173,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282421678198437e-05,
+      "loss": 4.168,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281583083447385e-05,
+      "loss": 4.167,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280744488696333e-05,
+      "loss": 4.1706,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279905893945281e-05,
+      "loss": 4.1683,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279067299194229e-05,
+      "loss": 4.1792,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782303423235496e-05,
+      "loss": 4.1832,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773917475724976e-05,
+      "loss": 4.1774,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765531528214456e-05,
+      "loss": 4.1761,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757145580703936e-05,
+      "loss": 4.1606,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748759633193416e-05,
+      "loss": 4.1689,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274039006448663e-05,
+      "loss": 4.1713,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273200411697611e-05,
+      "loss": 4.182,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272361816946559e-05,
+      "loss": 4.1528,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271523222195507e-05,
+      "loss": 4.172,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706846274444545e-05,
+      "loss": 4.1732,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269847670573776e-05,
+      "loss": 4.1707,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269010713703097e-05,
+      "loss": 4.165,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268172118952045e-05,
+      "loss": 4.1721,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267333524200993e-05,
+      "loss": 4.1637,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266494929449941e-05,
+      "loss": 4.1763,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265656334698889e-05,
+      "loss": 4.1725,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648193778282105e-05,
+      "loss": 4.1771,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639807830771585e-05,
+      "loss": 4.1568,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631421883261065e-05,
+      "loss": 4.1735,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623035935750545e-05,
+      "loss": 4.1771,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261464998824002e-05,
+      "loss": 4.1642,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606280419533234e-05,
+      "loss": 4.1681,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597894472022714e-05,
+      "loss": 4.1755,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.169,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.1826,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.1631,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1698,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255598111327384e-05,
+      "loss": 4.1641,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254759516576332e-05,
+      "loss": 4.168,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25392092182528e-05,
+      "loss": 4.1633,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253082327074228e-05,
+      "loss": 4.1591,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252243732323176e-05,
+      "loss": 4.1638,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251405137572124e-05,
+      "loss": 4.1549,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250568180701445e-05,
+      "loss": 4.1732,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.160346031188965,
+      "eval_runtime": 294.5727,
+      "eval_samples_per_second": 1295.405,
+      "eval_steps_per_second": 40.482,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249729585950393e-05,
+      "loss": 4.1725,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248890991199341e-05,
+      "loss": 4.1574,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248052396448289e-05,
+      "loss": 4.1687,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247213801697237e-05,
+      "loss": 4.1646,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246375206946185e-05,
+      "loss": 4.1741,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245536612195133e-05,
+      "loss": 4.161,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244698017444081e-05,
+      "loss": 4.1627,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243859422693029e-05,
+      "loss": 4.151,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243020827941978e-05,
+      "loss": 4.1606,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242182233190926e-05,
+      "loss": 4.1618,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241343638439873e-05,
+      "loss": 4.1646,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240505043688821e-05,
+      "loss": 4.16,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239668086818143e-05,
+      "loss": 4.1587,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238829492067091e-05,
+      "loss": 4.1604,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237990897316038e-05,
+      "loss": 4.1584,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237152302564986e-05,
+      "loss": 4.1459,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236313707813934e-05,
+      "loss": 4.1572,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235475113062882e-05,
+      "loss": 4.1584,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23463651831183e-05,
+      "loss": 4.157,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233797923560778e-05,
+      "loss": 4.1682,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329609666900996e-05,
+      "loss": 4.1646,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.1685,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.1508,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304451824369436e-05,
+      "loss": 4.1632,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1572,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287679929348396e-05,
+      "loss": 4.1623,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279293981837876e-05,
+      "loss": 4.1525,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270908034327356e-05,
+      "loss": 4.154,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262522086816836e-05,
+      "loss": 4.1515,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254152518110045e-05,
+      "loss": 4.1543,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245766570599525e-05,
+      "loss": 4.1554,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237397001892734e-05,
+      "loss": 4.1625,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229011054382214e-05,
+      "loss": 4.1609,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22206251068717e-05,
+      "loss": 4.159,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221223915936118e-05,
+      "loss": 4.1531,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220385321185066e-05,
+      "loss": 4.157,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219546726434014e-05,
+      "loss": 4.1603,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218709769563335e-05,
+      "loss": 4.1458,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217871174812283e-05,
+      "loss": 4.1402,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217032580061231e-05,
+      "loss": 4.141,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216193985310179e-05,
+      "loss": 4.1484,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215355390559126e-05,
+      "loss": 4.1541,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214516795808074e-05,
+      "loss": 4.151,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213678201057022e-05,
+      "loss": 4.1454,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21283960630597e-05,
+      "loss": 4.1514,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212002649435292e-05,
+      "loss": 4.1479,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21116405468424e-05,
+      "loss": 4.152,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1433,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094885030625094e-05,
+      "loss": 4.1247,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208649908311457e-05,
+      "loss": 4.1681,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207811313560405e-05,
+      "loss": 4.1508,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206972718809353e-05,
+      "loss": 4.1593,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206134124058301e-05,
+      "loss": 4.1469,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205295529307249e-05,
+      "loss": 4.149,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204456934556197e-05,
+      "loss": 4.1416,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1451,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.1471,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.1416,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1642,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1423,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199428641810631e-05,
+      "loss": 4.1315,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1574,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197751452308527e-05,
+      "loss": 4.1272,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196912857557475e-05,
+      "loss": 4.1291,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196074262806423e-05,
+      "loss": 4.153,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195235668055371e-05,
+      "loss": 4.152,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194397073304319e-05,
+      "loss": 4.1346,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193558478553267e-05,
+      "loss": 4.1431,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192723159562961e-05,
+      "loss": 4.1215,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191884564811909e-05,
+      "loss": 4.1388,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191045970060857e-05,
+      "loss": 4.1423,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.1388,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.1441,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1457,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876932289370226e-05,
+      "loss": 4.1566,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868546341859706e-05,
+      "loss": 4.1352,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860176773152915e-05,
+      "loss": 4.1392,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851790825642395e-05,
+      "loss": 4.1478,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843404878131875e-05,
+      "loss": 4.1337,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835018930621355e-05,
+      "loss": 4.144,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826632983110834e-05,
+      "loss": 4.1397,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818247035600314e-05,
+      "loss": 4.1362,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809861088089794e-05,
+      "loss": 4.1545,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801475140579274e-05,
+      "loss": 4.1264,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179310557187249e-05,
+      "loss": 4.1306,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178471962436197e-05,
+      "loss": 4.1417,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177635005565518e-05,
+      "loss": 4.144,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1347,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1414,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175119221312362e-05,
+      "loss": 4.1224,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17428062656131e-05,
+      "loss": 4.1336,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173442031810257e-05,
+      "loss": 4.1295,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172603437059205e-05,
+      "loss": 4.1366,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171764842308153e-05,
+      "loss": 4.1509,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170926247557101e-05,
+      "loss": 4.1383,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170087652806049e-05,
+      "loss": 4.1397,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169249058054998e-05,
+      "loss": 4.1291,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168410463303946e-05,
+      "loss": 4.1293,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167571868552894e-05,
+      "loss": 4.1293,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166736549562588e-05,
+      "loss": 4.1463,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165897954811536e-05,
+      "loss": 4.1374,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165059360060484e-05,
+      "loss": 4.1411,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164220765309432e-05,
+      "loss": 4.129,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1417,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625452136877006e-05,
+      "loss": 4.1259,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617066189366486e-05,
+      "loss": 4.1347,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608680241855966e-05,
+      "loss": 4.1195,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600294294345446e-05,
+      "loss": 4.1372,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159192472563866e-05,
+      "loss": 4.1258,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1335,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157515283061762e-05,
+      "loss": 4.1238,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.1331,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155838093559658e-05,
+      "loss": 4.1264,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154999498808606e-05,
+      "loss": 4.1261,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154160904057554e-05,
+      "loss": 4.1428,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153323947186875e-05,
+      "loss": 4.1429,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152485352435823e-05,
+      "loss": 4.1387,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151646757684771e-05,
+      "loss": 4.1339,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150808162933719e-05,
+      "loss": 4.1224,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14997120606304e-05,
+      "loss": 4.129,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149132611311988e-05,
+      "loss": 4.133,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482956544413095e-05,
+      "loss": 4.1379,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474570596902575e-05,
+      "loss": 4.1168,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.133,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457798701881535e-05,
+      "loss": 4.1324,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449429133174744e-05,
+      "loss": 4.1316,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441043185664224e-05,
+      "loss": 4.1273,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1301,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424271290643184e-05,
+      "loss": 4.1242,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.1372,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407499395622144e-05,
+      "loss": 4.1354,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399113448111624e-05,
+      "loss": 4.134,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390727500601104e-05,
+      "loss": 4.1169,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138237431069805e-05,
+      "loss": 4.1372,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137398836318753e-05,
+      "loss": 4.1355,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.1237,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1316,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1323,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134044457314545e-05,
+      "loss": 4.1301,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133205862563493e-05,
+      "loss": 4.1439,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132367267812441e-05,
+      "loss": 4.1231,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131528673061388e-05,
+      "loss": 4.1323,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1281,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129853121439658e-05,
+      "loss": 4.1285,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129014526688606e-05,
+      "loss": 4.1212,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128175931937554e-05,
+      "loss": 4.1186,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127337337186502e-05,
+      "loss": 4.1242,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649874243545e-05,
+      "loss": 4.1185,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125663423445144e-05,
+      "loss": 4.1335,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.130283355712891,
+      "eval_runtime": 300.7639,
+      "eval_samples_per_second": 1268.739,
+      "eval_steps_per_second": 39.649,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124824828694092e-05,
+      "loss": 4.13,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12398623394304e-05,
+      "loss": 4.1189,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123147639191988e-05,
+      "loss": 4.1292,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1223090444409356e-05,
+      "loss": 4.1283,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1214704496898835e-05,
+      "loss": 4.1348,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206318549388315e-05,
+      "loss": 4.1239,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197932601877795e-05,
+      "loss": 4.1302,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189546654367275e-05,
+      "loss": 4.1149,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118117708566049e-05,
+      "loss": 4.117,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117280751695371e-05,
+      "loss": 4.1305,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116442156944318e-05,
+      "loss": 4.1253,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115603562193266e-05,
+      "loss": 4.1197,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114764967442214e-05,
+      "loss": 4.129,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113926372691162e-05,
+      "loss": 4.1172,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11308777794011e-05,
+      "loss": 4.1155,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112249183189058e-05,
+      "loss": 4.1101,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111410588438006e-05,
+      "loss": 4.1183,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110573631567327e-05,
+      "loss": 4.1233,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109735036816275e-05,
+      "loss": 4.1147,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108896442065223e-05,
+      "loss": 4.1334,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108057847314171e-05,
+      "loss": 4.1267,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072192525631196e-05,
+      "loss": 4.1303,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063822956924405e-05,
+      "loss": 4.1164,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055437009413885e-05,
+      "loss": 4.1212,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1047051061903365e-05,
+      "loss": 4.1259,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038665114392845e-05,
+      "loss": 4.1232,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1147,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1178,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1013507271861285e-05,
+      "loss": 4.1116,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005137703154494e-05,
+      "loss": 4.117,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996751755643974e-05,
+      "loss": 4.1192,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988365808133454e-05,
+      "loss": 4.1241,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0979979860622933e-05,
+      "loss": 4.1254,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0971593913112413e-05,
+      "loss": 4.1241,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096320796560189e-05,
+      "loss": 4.1185,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095483839689511e-05,
+      "loss": 4.1177,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094646882818832e-05,
+      "loss": 4.1247,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09380828806778e-05,
+      "loss": 4.1086,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092969693316728e-05,
+      "loss": 4.1074,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092131098565676e-05,
+      "loss": 4.1027,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091292503814624e-05,
+      "loss": 4.1116,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090453909063572e-05,
+      "loss": 4.113,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089615314312519e-05,
+      "loss": 4.1164,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088776719561467e-05,
+      "loss": 4.1122,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087938124810415e-05,
+      "loss": 4.1125,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087099530059363e-05,
+      "loss": 4.1137,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086260935308311e-05,
+      "loss": 4.1185,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085423978437633e-05,
+      "loss": 4.1064,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084585383686581e-05,
+      "loss": 4.0905,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837484268159016e-05,
+      "loss": 4.1263,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829098320648496e-05,
+      "loss": 4.1171,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0820712373137976e-05,
+      "loss": 4.1248,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0812326425627456e-05,
+      "loss": 4.1098,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0803940478116936e-05,
+      "loss": 4.1136,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0795554530606416e-05,
+      "loss": 4.1068,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787168583095896e-05,
+      "loss": 4.1085,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778782635585376e-05,
+      "loss": 4.1159,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770413066878585e-05,
+      "loss": 4.1004,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762027119368065e-05,
+      "loss": 4.131,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075364117185755e-05,
+      "loss": 4.1114,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074525522434703e-05,
+      "loss": 4.0914,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073686927683651e-05,
+      "loss": 4.1203,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072849970812972e-05,
+      "loss": 4.0946,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07201137606192e-05,
+      "loss": 4.0932,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071172781310868e-05,
+      "loss": 4.1186,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.1119,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069495591808764e-05,
+      "loss": 4.0999,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068656997057712e-05,
+      "loss": 4.1076,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067820040187033e-05,
+      "loss": 4.0912,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066981445435981e-05,
+      "loss": 4.1017,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066142850684929e-05,
+      "loss": 4.1073,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065304255933877e-05,
+      "loss": 4.1036,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644672990631985e-05,
+      "loss": 4.1104,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636287043121465e-05,
+      "loss": 4.1092,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627901095610945e-05,
+      "loss": 4.12,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619515148100425e-05,
+      "loss": 4.1008,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611129200589905e-05,
+      "loss": 4.1054,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602759631883114e-05,
+      "loss": 4.1102,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594373684372594e-05,
+      "loss": 4.0987,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585987736862074e-05,
+      "loss": 4.1112,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577601789351554e-05,
+      "loss": 4.1068,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056921584184103e-05,
+      "loss": 4.0965,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056082989433051e-05,
+      "loss": 4.1192,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055244394681999e-05,
+      "loss": 4.0926,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05440743781132e-05,
+      "loss": 4.0941,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053568843060268e-05,
+      "loss": 4.1098,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052730248309216e-05,
+      "loss": 4.1072,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051891653558164e-05,
+      "loss": 4.0997,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051053058807112e-05,
+      "loss": 4.1106,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050216101936433e-05,
+      "loss": 4.0873,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049377507185381e-05,
+      "loss": 4.0967,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048538912434329e-05,
+      "loss": 4.0995,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047700317683277e-05,
+      "loss": 4.0999,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046861722932225e-05,
+      "loss": 4.1168,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046024766061546e-05,
+      "loss": 4.1052,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.1077,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044347576559443e-05,
+      "loss": 4.0948,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043508981808391e-05,
+      "loss": 4.0957,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042670387057339e-05,
+      "loss": 4.0975,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041831792306287e-05,
+      "loss": 4.1084,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040993197555235e-05,
+      "loss": 4.1035,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040154602804183e-05,
+      "loss": 4.1081,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0393176459335036e-05,
+      "loss": 4.0969,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0384790511824516e-05,
+      "loss": 4.1032,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0376404564313996e-05,
+      "loss": 4.0979,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368034995607205e-05,
+      "loss": 4.0987,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359649048096685e-05,
+      "loss": 4.0856,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351263100586165e-05,
+      "loss": 4.1066,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342877153075645e-05,
+      "loss": 4.0908,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334491205565125e-05,
+      "loss": 4.0981,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326105258054605e-05,
+      "loss": 4.089,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031773568934782e-05,
+      "loss": 4.0975,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03093497418373e-05,
+      "loss": 4.0952,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030096379432678e-05,
+      "loss": 4.0932,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029257784681626e-05,
+      "loss": 4.1045,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028420827810947e-05,
+      "loss": 4.1122,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027582233059895e-05,
+      "loss": 4.1097,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026743638308843e-05,
+      "loss": 4.0983,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025905043557791e-05,
+      "loss": 4.0911,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025066448806738e-05,
+      "loss": 4.0936,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024227854055686e-05,
+      "loss": 4.1006,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023390897185008e-05,
+      "loss": 4.1033,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022552302433956e-05,
+      "loss": 4.0838,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021713707682904e-05,
+      "loss": 4.1011,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020875112931852e-05,
+      "loss": 4.1022,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200365181808e-05,
+      "loss": 4.0952,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0191995613101214e-05,
+      "loss": 4.0946,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018360966559069e-05,
+      "loss": 4.101,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017522371808017e-05,
+      "loss": 4.0902,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016683777056965e-05,
+      "loss": 4.1028,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015845182305913e-05,
+      "loss": 4.0999,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015008225435234e-05,
+      "loss": 4.1049,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0141696306841817e-05,
+      "loss": 4.0823,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133310359331297e-05,
+      "loss": 4.1052,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012492441182078e-05,
+      "loss": 4.102,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.094,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010816889560347e-05,
+      "loss": 4.0991,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009978294809295e-05,
+      "loss": 4.0959,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009139700058243e-05,
+      "loss": 4.0988,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008301105307191e-05,
+      "loss": 4.1128,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007462510556139e-05,
+      "loss": 4.0869,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00662555368546e-05,
+      "loss": 4.1011,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005786958934408e-05,
+      "loss": 4.0957,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004948364183356e-05,
+      "loss": 4.097,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004109769432304e-05,
+      "loss": 4.0882,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003271174681252e-05,
+      "loss": 4.0835,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002434217810574e-05,
+      "loss": 4.0934,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001595623059522e-05,
+      "loss": 4.0854,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00075702830847e-05,
+      "loss": 4.1006,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.106624126434326,
+      "eval_runtime": 297.2422,
+      "eval_samples_per_second": 1283.771,
+      "eval_steps_per_second": 40.119,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999918433557418e-05,
+      "loss": 4.0956,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999079838806366e-05,
+      "loss": 4.0894,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998241244055314e-05,
+      "loss": 4.096,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997402649304262e-05,
+      "loss": 4.0955,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99656405455321e-05,
+      "loss": 4.1006,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957254598021577e-05,
+      "loss": 4.0933,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994886865051105e-05,
+      "loss": 4.0983,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994048270300053e-05,
+      "loss": 4.0825,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993209675549001e-05,
+      "loss": 4.0852,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992371080797949e-05,
+      "loss": 4.0994,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991532486046897e-05,
+      "loss": 4.0927,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990693891295845e-05,
+      "loss": 4.0866,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989855296544793e-05,
+      "loss": 4.0975,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890167017937416e-05,
+      "loss": 4.0837,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881781070426896e-05,
+      "loss": 4.0857,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873395122916376e-05,
+      "loss": 4.0733,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865009175405856e-05,
+      "loss": 4.0886,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856623227895336e-05,
+      "loss": 4.0954,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848253659188545e-05,
+      "loss": 4.0813,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839867711678025e-05,
+      "loss": 4.1009,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.0963,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823095816656985e-05,
+      "loss": 4.0964,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981470986914646e-05,
+      "loss": 4.0885,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806340300439674e-05,
+      "loss": 4.092,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.093,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789568405418634e-05,
+      "loss": 4.0898,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781182457908114e-05,
+      "loss": 4.0832,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772796510397594e-05,
+      "loss": 4.0896,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976442694169081e-05,
+      "loss": 4.0802,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975604099418029e-05,
+      "loss": 4.0844,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974765504666976e-05,
+      "loss": 4.0881,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973926909915924e-05,
+      "loss": 4.0907,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973088315164872e-05,
+      "loss": 4.0957,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97224972041382e-05,
+      "loss": 4.0966,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971412763543141e-05,
+      "loss": 4.0816,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970574168792089e-05,
+      "loss": 4.0914,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969735574041037e-05,
+      "loss": 4.0943,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968896979289985e-05,
+      "loss": 4.0754,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968058384538934e-05,
+      "loss": 4.0789,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967219789787882e-05,
+      "loss": 4.0761,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.0754,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965544238166151e-05,
+      "loss": 4.084,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964705643415099e-05,
+      "loss": 4.0866,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963867048664047e-05,
+      "loss": 4.0805,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963028453912995e-05,
+      "loss": 4.083,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621914970423156e-05,
+      "loss": 4.0818,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613529022912636e-05,
+      "loss": 4.0908,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605143075402116e-05,
+      "loss": 4.0735,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9596757127891596e-05,
+      "loss": 4.0625,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588371180381076e-05,
+      "loss": 4.094,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9579985232870556e-05,
+      "loss": 4.0869,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571599285360036e-05,
+      "loss": 4.0927,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956321333784952e-05,
+      "loss": 4.0789,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554827390339e-05,
+      "loss": 4.0818,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954645782163221e-05,
+      "loss": 4.0808,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953807187412169e-05,
+      "loss": 4.0788,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952968592661117e-05,
+      "loss": 4.0864,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952131635790438e-05,
+      "loss": 4.068,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951293041039386e-05,
+      "loss": 4.1032,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950454446288334e-05,
+      "loss": 4.0783,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949615851537282e-05,
+      "loss": 4.0604,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487772567862294e-05,
+      "loss": 4.0923,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479386620351774e-05,
+      "loss": 4.0709,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947100067284126e-05,
+      "loss": 4.0603,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946261472533074e-05,
+      "loss": 4.0868,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945424515662395e-05,
+      "loss": 4.0842,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445875587917166e-05,
+      "loss": 4.0724,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437489640406646e-05,
+      "loss": 4.0762,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429103692896126e-05,
+      "loss": 4.0602,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94207177453856e-05,
+      "loss": 4.0692,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941233179787508e-05,
+      "loss": 4.0816,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940394585036456e-05,
+      "loss": 4.0739,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939555990285404e-05,
+      "loss": 4.0778,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938719033414725e-05,
+      "loss": 4.0763,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937880438663673e-05,
+      "loss": 4.0924,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370418439126214e-05,
+      "loss": 4.0753,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362032491615694e-05,
+      "loss": 4.0752,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0789,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9345260596594654e-05,
+      "loss": 4.0721,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9336874649084134e-05,
+      "loss": 4.0819,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9328488701573614e-05,
+      "loss": 4.0764,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320102754063094e-05,
+      "loss": 4.0707,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311716806552574e-05,
+      "loss": 4.0878,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930336361664951e-05,
+      "loss": 4.0631,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929497766913899e-05,
+      "loss": 4.0694,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928659172162847e-05,
+      "loss": 4.078,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927822215292168e-05,
+      "loss": 4.0786,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926983620541116e-05,
+      "loss": 4.0684,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926145025790065e-05,
+      "loss": 4.0814,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925306431039013e-05,
+      "loss": 4.0639,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924467836287961e-05,
+      "loss": 4.0691,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923629241536909e-05,
+      "loss": 4.0654,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922790646785857e-05,
+      "loss": 4.0716,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921952052034805e-05,
+      "loss": 4.088,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921113457283753e-05,
+      "loss": 4.0752,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920274862532701e-05,
+      "loss": 4.0775,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919436267781648e-05,
+      "loss": 4.0684,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918597673030596e-05,
+      "loss": 4.066,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917759078279544e-05,
+      "loss": 4.0682,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916920483528492e-05,
+      "loss": 4.0807,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91608188877744e-05,
+      "loss": 4.0757,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152449319067617e-05,
+      "loss": 4.074,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144063371557096e-05,
+      "loss": 4.0762,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9135677424046576e-05,
+      "loss": 4.0712,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127291476536056e-05,
+      "loss": 4.0707,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118905529025536e-05,
+      "loss": 4.0689,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9110519581515016e-05,
+      "loss": 4.0605,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102133634004496e-05,
+      "loss": 4.0758,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093747686493976e-05,
+      "loss": 4.0642,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085394496590914e-05,
+      "loss": 4.069,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077008549080394e-05,
+      "loss": 4.066,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068622601569874e-05,
+      "loss": 4.0635,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060236654059354e-05,
+      "loss": 4.0674,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051850706548834e-05,
+      "loss": 4.0649,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904348113784205e-05,
+      "loss": 4.0819,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903509519033153e-05,
+      "loss": 4.0794,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902670924282101e-05,
+      "loss": 4.079,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901832329531049e-05,
+      "loss": 4.0695,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900993734779997e-05,
+      "loss": 4.067,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900155140028945e-05,
+      "loss": 4.0653,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899316545277893e-05,
+      "loss": 4.0727,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898479588407214e-05,
+      "loss": 4.0773,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897640993656162e-05,
+      "loss": 4.0585,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89680239890511e-05,
+      "loss": 4.0691,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895963804154058e-05,
+      "loss": 4.0717,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895126847283379e-05,
+      "loss": 4.0701,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894288252532327e-05,
+      "loss": 4.0658,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934496577812755e-05,
+      "loss": 4.0712,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926110630302235e-05,
+      "loss": 4.0616,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917724682791715e-05,
+      "loss": 4.0768,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.0673,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8900969166574404e-05,
+      "loss": 4.0818,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892583219063884e-05,
+      "loss": 4.0572,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884197271553364e-05,
+      "loss": 4.0747,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875811324042843e-05,
+      "loss": 4.0788,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886744175533605e-05,
+      "loss": 4.0668,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885905580782553e-05,
+      "loss": 4.0684,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885066986031501e-05,
+      "loss": 4.0701,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884228391280449e-05,
+      "loss": 4.069,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883389796529397e-05,
+      "loss": 4.0832,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882552839658719e-05,
+      "loss": 4.0632,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881714244907667e-05,
+      "loss": 4.0736,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880875650156615e-05,
+      "loss": 4.0703,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880037055405562e-05,
+      "loss": 4.0682,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87919846065451e-05,
+      "loss": 4.0628,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878361503783832e-05,
+      "loss": 4.0579,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877522909032779e-05,
+      "loss": 4.0634,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876684314281727e-05,
+      "loss": 4.0642,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875845719530675e-05,
+      "loss": 4.0695,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.087048053741455,
+      "eval_runtime": 290.952,
+      "eval_samples_per_second": 1311.526,
+      "eval_steps_per_second": 40.986,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875007124779623e-05,
+      "loss": 4.0661,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874170167908944e-05,
+      "loss": 4.0625,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733315731578926e-05,
+      "loss": 4.0649,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8724929784068406e-05,
+      "loss": 4.0724,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8716543836557886e-05,
+      "loss": 4.0773,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708174267851095e-05,
+      "loss": 4.0668,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8699788320340575e-05,
+      "loss": 4.066,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8691402372830055e-05,
+      "loss": 4.0571,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683032804123264e-05,
+      "loss": 4.0595,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674646856612744e-05,
+      "loss": 4.0689,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666260909102224e-05,
+      "loss": 4.0688,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657874961591704e-05,
+      "loss": 4.0615,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649489014081184e-05,
+      "loss": 4.0692,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641103066570664e-05,
+      "loss": 4.0582,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632717119060144e-05,
+      "loss": 4.0595,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862434755035336e-05,
+      "loss": 4.045,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861596160284284e-05,
+      "loss": 4.0667,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860757565533232e-05,
+      "loss": 4.0677,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85991897078218e-05,
+      "loss": 4.053,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859080376031128e-05,
+      "loss": 4.071,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858241781280076e-05,
+      "loss": 4.0691,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857403186529024e-05,
+      "loss": 4.0711,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856564591777972e-05,
+      "loss": 4.0631,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855727634907293e-05,
+      "loss": 4.0651,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854889040156241e-05,
+      "loss": 4.0644,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854050445405189e-05,
+      "loss": 4.0692,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853211850654137e-05,
+      "loss": 4.0506,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852373255903085e-05,
+      "loss": 4.0637,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851534661152033e-05,
+      "loss": 4.0524,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8506977042813544e-05,
+      "loss": 4.0623,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8498591095303024e-05,
+      "loss": 4.0638,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490205147792504e-05,
+      "loss": 4.0656,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848181920028198e-05,
+      "loss": 4.0615,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847343325277146e-05,
+      "loss": 4.0744,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846504730526094e-05,
+      "loss": 4.0567,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845667773655415e-05,
+      "loss": 4.0626,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448291789043626e-05,
+      "loss": 4.0677,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439905841533106e-05,
+      "loss": 4.0483,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431519894022586e-05,
+      "loss": 4.052,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423133946512066e-05,
+      "loss": 4.0531,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414747999001546e-05,
+      "loss": 4.0483,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840636205149103e-05,
+      "loss": 4.0573,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839797610398051e-05,
+      "loss": 4.064,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838960653527372e-05,
+      "loss": 4.0501,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83812205877632e-05,
+      "loss": 4.0568,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837283464025268e-05,
+      "loss": 4.0539,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836444869274216e-05,
+      "loss": 4.0606,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835607912403537e-05,
+      "loss": 4.0482,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834770955532858e-05,
+      "loss": 4.0388,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833932360781806e-05,
+      "loss": 4.0683,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833093766030754e-05,
+      "loss": 4.0611,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832255171279702e-05,
+      "loss": 4.0718,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83141657652865e-05,
+      "loss": 4.0502,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305779817775986e-05,
+      "loss": 4.0551,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297410249069195e-05,
+      "loss": 4.0544,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8289024301558675e-05,
+      "loss": 4.0489,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8280638354048155e-05,
+      "loss": 4.0647,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272252406537635e-05,
+      "loss": 4.0384,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263866459027115e-05,
+      "loss": 4.0804,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255496890320324e-05,
+      "loss": 4.0564,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247127321613534e-05,
+      "loss": 4.0324,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238741374103013e-05,
+      "loss": 4.0625,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230355426592493e-05,
+      "loss": 4.0488,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822196947908197e-05,
+      "loss": 4.0358,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821358353157145e-05,
+      "loss": 4.0584,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820519758406093e-05,
+      "loss": 4.0621,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819681163655042e-05,
+      "loss": 4.0458,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81884256890399e-05,
+      "loss": 4.0529,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818003974152938e-05,
+      "loss": 4.0312,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817167017282259e-05,
+      "loss": 4.0451,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816328422531207e-05,
+      "loss": 4.0525,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815489827780155e-05,
+      "loss": 4.0467,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814651233029103e-05,
+      "loss": 4.0588,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813812638278051e-05,
+      "loss": 4.0476,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812974043526999e-05,
+      "loss": 4.0708,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812135448775946e-05,
+      "loss": 4.0492,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811298491905268e-05,
+      "loss": 4.053,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810459897154216e-05,
+      "loss": 4.0546,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809621302403164e-05,
+      "loss": 4.0466,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808782707652112e-05,
+      "loss": 4.0551,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80794411290106e-05,
+      "loss": 4.0474,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807105518150008e-05,
+      "loss": 4.0516,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806266923398956e-05,
+      "loss": 4.0583,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805428328647904e-05,
+      "loss": 4.0474,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804591371777225e-05,
+      "loss": 4.0361,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037527770261727e-05,
+      "loss": 4.0511,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029158201554936e-05,
+      "loss": 4.0572,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8020772254044416e-05,
+      "loss": 4.0433,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012386306533896e-05,
+      "loss": 4.0537,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004000359023376e-05,
+      "loss": 4.0387,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7995614411512856e-05,
+      "loss": 4.045,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798724484280607e-05,
+      "loss": 4.0405,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797885889529555e-05,
+      "loss": 4.0452,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797047294778503e-05,
+      "loss": 4.06,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796208700027451e-05,
+      "loss": 4.0525,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795370105276399e-05,
+      "loss": 4.0538,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79453314840572e-05,
+      "loss": 4.0462,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793694553654668e-05,
+      "loss": 4.0408,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792855958903616e-05,
+      "loss": 4.0444,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792017364152564e-05,
+      "loss": 4.0534,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791178769401512e-05,
+      "loss": 4.0518,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79034017465046e-05,
+      "loss": 4.0523,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789501579899408e-05,
+      "loss": 4.0502,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788662985148356e-05,
+      "loss": 4.0451,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787824390397304e-05,
+      "loss": 4.0498,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7869874335266256e-05,
+      "loss": 4.0438,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7861488387755736e-05,
+      "loss": 4.0391,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853102440245216e-05,
+      "loss": 4.0486,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844716492734696e-05,
+      "loss": 4.0413,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836346924027905e-05,
+      "loss": 4.043,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7827960976517385e-05,
+      "loss": 4.0384,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7819575029006865e-05,
+      "loss": 4.0387,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7811189081496345e-05,
+      "loss": 4.0464,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802803133985825e-05,
+      "loss": 4.0394,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794433565279034e-05,
+      "loss": 4.0556,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778606399657225e-05,
+      "loss": 4.0543,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777767804906173e-05,
+      "loss": 4.058,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776930848035494e-05,
+      "loss": 4.0466,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776092253284442e-05,
+      "loss": 4.044,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77525365853339e-05,
+      "loss": 4.0386,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774415063782338e-05,
+      "loss": 4.0545,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773576469031286e-05,
+      "loss": 4.0505,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772737874280234e-05,
+      "loss": 4.0341,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771899279529182e-05,
+      "loss": 4.0436,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77106068477813e-05,
+      "loss": 4.0463,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770222090027077e-05,
+      "loss": 4.0494,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769385133156399e-05,
+      "loss": 4.0427,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768546538405347e-05,
+      "loss": 4.0444,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767707943654295e-05,
+      "loss": 4.0388,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766869348903243e-05,
+      "loss": 4.0543,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766030754152191e-05,
+      "loss": 4.0442,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765192159401139e-05,
+      "loss": 4.0551,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764353564650087e-05,
+      "loss": 4.0354,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763514969899035e-05,
+      "loss": 4.0495,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7626780130283556e-05,
+      "loss": 4.055,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7618394182773036e-05,
+      "loss": 4.0406,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7610024614066245e-05,
+      "loss": 4.048,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7601638666555725e-05,
+      "loss": 4.046,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7593252719045205e-05,
+      "loss": 4.0483,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7584866771534685e-05,
+      "loss": 4.0552,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7576480824024165e-05,
+      "loss": 4.0449,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756811125531738e-05,
+      "loss": 4.0464,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755972530780686e-05,
+      "loss": 4.0485,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755133936029634e-05,
+      "loss": 4.0451,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754295341278582e-05,
+      "loss": 4.0391,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75345674652753e-05,
+      "loss": 4.0351,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752619789656851e-05,
+      "loss": 4.0402,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751782832786172e-05,
+      "loss": 4.041,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75094423803512e-05,
+      "loss": 4.0465,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.071537017822266,
+      "eval_runtime": 290.8862,
+      "eval_samples_per_second": 1311.822,
+      "eval_steps_per_second": 40.995,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750105643284068e-05,
+      "loss": 4.047,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749267048533016e-05,
+      "loss": 4.0377,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748428453781964e-05,
+      "loss": 4.0415,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747589859030912e-05,
+      "loss": 4.0506,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467512642798605e-05,
+      "loss": 4.0532,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459126695288085e-05,
+      "loss": 4.0437,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450740747777565e-05,
+      "loss": 4.0465,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7442354800267045e-05,
+      "loss": 4.0367,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433968852756525e-05,
+      "loss": 4.0352,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425582905246005e-05,
+      "loss": 4.0426,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417196957735485e-05,
+      "loss": 4.0501,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740881101022496e-05,
+      "loss": 4.0398,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7400441441518174e-05,
+      "loss": 4.0458,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392055494007654e-05,
+      "loss": 4.0345,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738366954649713e-05,
+      "loss": 4.0389,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737528359898661e-05,
+      "loss": 4.0201,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736691403027982e-05,
+      "loss": 4.045,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73585280827693e-05,
+      "loss": 4.043,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735014213525878e-05,
+      "loss": 4.0307,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7341772566552e-05,
+      "loss": 4.0492,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733338661904148e-05,
+      "loss": 4.0493,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732500067153096e-05,
+      "loss": 4.0478,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731661472402043e-05,
+      "loss": 4.0388,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730822877650991e-05,
+      "loss": 4.0423,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729984282899939e-05,
+      "loss": 4.0417,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729145688148887e-05,
+      "loss": 4.0451,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728307093397835e-05,
+      "loss": 4.0293,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727468498646783e-05,
+      "loss": 4.041,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726629903895731e-05,
+      "loss": 4.0342,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725791309144679e-05,
+      "loss": 4.0342,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724952714393627e-05,
+      "loss": 4.0418,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724114119642576e-05,
+      "loss": 4.0436,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723275524891524e-05,
+      "loss": 4.0411,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722436930140472e-05,
+      "loss": 4.0487,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72159833538942e-05,
+      "loss": 4.0355,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720761378518741e-05,
+      "loss": 4.0409,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719922783767689e-05,
+      "loss": 4.0469,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719084189016637e-05,
+      "loss": 4.0249,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718245594265585e-05,
+      "loss": 4.0279,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717406999514532e-05,
+      "loss": 4.0318,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71656840476348e-05,
+      "loss": 4.0218,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157314478928016e-05,
+      "loss": 4.0366,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148928531417496e-05,
+      "loss": 4.0447,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140542583906976e-05,
+      "loss": 4.0285,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132156636396456e-05,
+      "loss": 4.0364,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123770688885936e-05,
+      "loss": 4.0364,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115384741375416e-05,
+      "loss": 4.0375,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7106998793864896e-05,
+      "loss": 4.0259,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7098629225158105e-05,
+      "loss": 4.0157,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7090243277647585e-05,
+      "loss": 4.0432,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7081857330137065e-05,
+      "loss": 4.0433,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7073471382626545e-05,
+      "loss": 4.0522,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7065085435116025e-05,
+      "loss": 4.0292,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7056715866409234e-05,
+      "loss": 4.0321,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048329918898714e-05,
+      "loss": 4.0328,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7039943971388194e-05,
+      "loss": 4.0264,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7031558023877674e-05,
+      "loss": 4.0453,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702317207636716e-05,
+      "loss": 4.0143,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701478612885664e-05,
+      "loss": 4.0596,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700641656014985e-05,
+      "loss": 4.0367,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699803061263933e-05,
+      "loss": 4.014,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698964466512881e-05,
+      "loss": 4.04,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698125871761829e-05,
+      "loss": 4.0276,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697287277010777e-05,
+      "loss": 4.0132,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696448682259725e-05,
+      "loss": 4.0385,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695611725389046e-05,
+      "loss": 4.0389,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694773130637994e-05,
+      "loss": 4.0231,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693934535886942e-05,
+      "loss": 4.0369,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693097579016263e-05,
+      "loss": 4.0096,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922589842652114e-05,
+      "loss": 4.0254,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6914203895141594e-05,
+      "loss": 4.027,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905817947631074e-05,
+      "loss": 4.031,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897432000120554e-05,
+      "loss": 4.0335,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889046052610034e-05,
+      "loss": 4.0285,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688066010509951e-05,
+      "loss": 4.0503,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687227415758899e-05,
+      "loss": 4.0299,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686388821007847e-05,
+      "loss": 4.0334,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685553502017541e-05,
+      "loss": 4.0289,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684714907266489e-05,
+      "loss": 4.0255,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683876312515437e-05,
+      "loss": 4.0346,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683037717764385e-05,
+      "loss": 4.0257,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682199123013333e-05,
+      "loss": 4.0304,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681360528262281e-05,
+      "loss": 4.0425,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680521933511229e-05,
+      "loss": 4.0265,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679683338760177e-05,
+      "loss": 4.0148,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678846381889498e-05,
+      "loss": 4.0284,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67800942501882e-05,
+      "loss": 4.0379,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677170830267768e-05,
+      "loss": 4.0184,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676332235516715e-05,
+      "loss": 4.0361,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675493640765663e-05,
+      "loss": 4.0237,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674655046014611e-05,
+      "loss": 4.0171,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738180891439326e-05,
+      "loss": 4.0222,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729794943928806e-05,
+      "loss": 4.025,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721408996418286e-05,
+      "loss": 4.044,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713023048907766e-05,
+      "loss": 4.0285,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704637101397246e-05,
+      "loss": 4.0317,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696251153886726e-05,
+      "loss": 4.0234,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687865206376206e-05,
+      "loss": 4.0193,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6679479258865685e-05,
+      "loss": 4.0221,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671093311355165e-05,
+      "loss": 4.0347,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662707363844645e-05,
+      "loss": 4.0341,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654321416334125e-05,
+      "loss": 4.0299,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6645935468823605e-05,
+      "loss": 4.0307,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6637565900116814e-05,
+      "loss": 4.0196,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629179952606294e-05,
+      "loss": 4.0307,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620794005095774e-05,
+      "loss": 4.0247,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661244081519272e-05,
+      "loss": 4.0182,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66040548676822e-05,
+      "loss": 4.0245,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659566892017168e-05,
+      "loss": 4.0229,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658728297266116e-05,
+      "loss": 4.0242,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657889702515064e-05,
+      "loss": 4.0175,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657051107764012e-05,
+      "loss": 4.018,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65621251301296e-05,
+      "loss": 4.0236,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655373918261908e-05,
+      "loss": 4.0263,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654535323510856e-05,
+      "loss": 4.0276,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653696728759804e-05,
+      "loss": 4.034,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652858134008751e-05,
+      "loss": 4.0401,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652019539257699e-05,
+      "loss": 4.0256,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651182582387021e-05,
+      "loss": 4.0252,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6503456255163424e-05,
+      "loss": 4.0181,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6495070307652904e-05,
+      "loss": 4.0308,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486684360142384e-05,
+      "loss": 4.0292,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478298412631864e-05,
+      "loss": 4.0186,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469912465121344e-05,
+      "loss": 4.0197,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646152651761082e-05,
+      "loss": 4.0309,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64531405701003e-05,
+      "loss": 4.0286,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644475462258978e-05,
+      "loss": 4.0204,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6436385053882986e-05,
+      "loss": 4.023,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6427999106372466e-05,
+      "loss": 4.0224,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6419613158861946e-05,
+      "loss": 4.0309,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6411227211351426e-05,
+      "loss": 4.0286,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640285764264464e-05,
+      "loss": 4.0347,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639447169513412e-05,
+      "loss": 4.0156,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63860857476236e-05,
+      "loss": 4.0303,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637769980011308e-05,
+      "loss": 4.034,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636933023140629e-05,
+      "loss": 4.0168,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636094428389577e-05,
+      "loss": 4.0253,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635255833638525e-05,
+      "loss": 4.028,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634417238887473e-05,
+      "loss": 4.031,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633578644136421e-05,
+      "loss": 4.0349,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632740049385369e-05,
+      "loss": 4.0229,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319047303950635e-05,
+      "loss": 4.0263,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310661356440115e-05,
+      "loss": 4.0259,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6302275408929595e-05,
+      "loss": 4.0216,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629390584022281e-05,
+      "loss": 4.0188,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6285519892712284e-05,
+      "loss": 4.0175,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6277133945201764e-05,
+      "loss": 4.0149,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6268747997691244e-05,
+      "loss": 4.0236,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6260362050180724e-05,
+      "loss": 4.0224,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.058211326599121,
+      "eval_runtime": 297.8122,
+      "eval_samples_per_second": 1281.314,
+      "eval_steps_per_second": 40.042,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251976102670204e-05,
+      "loss": 4.0236,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243590155159684e-05,
+      "loss": 4.0202,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623522058645289e-05,
+      "loss": 4.024,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622683463894237e-05,
+      "loss": 4.0274,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621844869143185e-05,
+      "loss": 4.0324,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621006274392133e-05,
+      "loss": 4.024,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620167679641081e-05,
+      "loss": 4.0268,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61932908489003e-05,
+      "loss": 4.0182,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618490490138978e-05,
+      "loss": 4.0131,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617651895387926e-05,
+      "loss": 4.0197,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616813300636874e-05,
+      "loss": 4.0334,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615974705885822e-05,
+      "loss": 4.0154,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615137749015143e-05,
+      "loss": 4.028,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614299154264091e-05,
+      "loss": 4.0123,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613460559513039e-05,
+      "loss": 4.0156,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612621964761987e-05,
+      "loss": 4.003,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611783370010935e-05,
+      "loss": 4.0248,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610944775259882e-05,
+      "loss": 4.0226,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61010618050883e-05,
+      "loss": 4.0151,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609267585757778e-05,
+      "loss": 4.0183,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608428991006727e-05,
+      "loss": 4.0338,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607590396255675e-05,
+      "loss": 4.0249,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606753439384996e-05,
+      "loss": 4.0227,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605914844633944e-05,
+      "loss": 4.0205,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605076249882892e-05,
+      "loss": 4.0282,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60423765513184e-05,
+      "loss": 4.0215,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603399060380788e-05,
+      "loss": 4.014,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602560465629736e-05,
+      "loss": 4.0186,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601721870878684e-05,
+      "loss": 4.0138,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600883276127632e-05,
+      "loss": 4.0106,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.60004468137658e-05,
+      "loss": 4.0241,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5992077245059006e-05,
+      "loss": 4.0173,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5983691297548486e-05,
+      "loss": 4.027,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5975305350037966e-05,
+      "loss": 4.0271,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5966919402527446e-05,
+      "loss": 4.0136,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595854983382066e-05,
+      "loss": 4.0223,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595016388631014e-05,
+      "loss": 4.0248,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594177793879962e-05,
+      "loss": 4.009,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593340837009283e-05,
+      "loss": 4.0133,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592502242258231e-05,
+      "loss": 4.007,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591663647507179e-05,
+      "loss": 4.0055,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590825052756127e-05,
+      "loss": 4.0175,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589986458005075e-05,
+      "loss": 4.0261,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589147863254023e-05,
+      "loss": 4.0093,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588309268502971e-05,
+      "loss": 4.0118,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874706737519184e-05,
+      "loss": 4.0177,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586632079000867e-05,
+      "loss": 4.0137,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585793484249815e-05,
+      "loss": 4.0095,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5849565273791366e-05,
+      "loss": 3.9976,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584117932628084e-05,
+      "loss": 4.0187,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583279337877032e-05,
+      "loss": 4.0258,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58244074312598e-05,
+      "loss": 4.0274,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581602148374928e-05,
+      "loss": 4.0111,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580765191504249e-05,
+      "loss": 4.0112,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5799282346335704e-05,
+      "loss": 4.0133,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5790896398825184e-05,
+      "loss": 4.0052,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578251045131466e-05,
+      "loss": 4.0247,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577412450380414e-05,
+      "loss": 3.9971,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5765738556293624e-05,
+      "loss": 4.0347,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5757352608783104e-05,
+      "loss": 4.0156,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5748966661272584e-05,
+      "loss": 3.9971,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5740580713762064e-05,
+      "loss": 4.0161,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573221114505527e-05,
+      "loss": 4.0093,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572382519754475e-05,
+      "loss": 3.9948,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571543925003423e-05,
+      "loss": 4.0151,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570705330252371e-05,
+      "loss": 4.0215,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569866735501319e-05,
+      "loss": 4.0036,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569028140750267e-05,
+      "loss": 4.0173,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568191183879588e-05,
+      "loss": 3.9905,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567352589128536e-05,
+      "loss": 4.0078,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566513994377484e-05,
+      "loss": 4.0079,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565675399626432e-05,
+      "loss": 4.0132,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564836804875381e-05,
+      "loss": 4.0093,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563999848004702e-05,
+      "loss": 4.0118,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56316125325365e-05,
+      "loss": 4.0292,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562324296382971e-05,
+      "loss": 4.0135,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561485701631919e-05,
+      "loss": 4.013,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5606471068808667e-05,
+      "loss": 4.0093,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5598085121298147e-05,
+      "loss": 4.0064,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589699173787627e-05,
+      "loss": 4.0169,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581313226277106e-05,
+      "loss": 4.0085,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5572943657570316e-05,
+      "loss": 4.0071,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5564557710059796e-05,
+      "loss": 4.0201,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5556171762549275e-05,
+      "loss": 4.0095,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547785815038755e-05,
+      "loss": 3.9918,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553939986752824e-05,
+      "loss": 4.0106,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553101392001772e-05,
+      "loss": 4.0147,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552264435131093e-05,
+      "loss": 4.002,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551425840380041e-05,
+      "loss": 4.0139,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550587245628989e-05,
+      "loss": 4.0066,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549748650877937e-05,
+      "loss": 3.9978,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489100561268844e-05,
+      "loss": 4.0081,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5480714613758324e-05,
+      "loss": 4.0019,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5472328666247804e-05,
+      "loss": 4.0226,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546395909754102e-05,
+      "loss": 4.0124,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545558952883423e-05,
+      "loss": 4.0164,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544720358132371e-05,
+      "loss": 4.0039,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438817633813196e-05,
+      "loss": 4.0003,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543043168630267e-05,
+      "loss": 4.0077,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542204573879215e-05,
+      "loss": 4.0148,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541365979128163e-05,
+      "loss": 4.0143,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540527384377111e-05,
+      "loss": 4.0091,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539688789626059e-05,
+      "loss": 4.0094,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53885183275538e-05,
+      "loss": 4.0006,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538013238004328e-05,
+      "loss": 4.015,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537174643253276e-05,
+      "loss": 4.0033,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536336048502224e-05,
+      "loss": 4.0066,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535497453751172e-05,
+      "loss": 4.0037,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53465885900012e-05,
+      "loss": 4.0032,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533820264249068e-05,
+      "loss": 4.0078,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5329816694980164e-05,
+      "loss": 3.9951,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321430747469644e-05,
+      "loss": 3.9976,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313044799959124e-05,
+      "loss": 4.008,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304658852448604e-05,
+      "loss": 4.0059,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296272904938084e-05,
+      "loss": 4.01,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528790333623129e-05,
+      "loss": 4.0145,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527951738872077e-05,
+      "loss": 4.0226,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527114782001398e-05,
+      "loss": 4.0049,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526276187250346e-05,
+      "loss": 4.0033,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525437592499294e-05,
+      "loss": 4.0036,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524598997748242e-05,
+      "loss": 4.0116,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52376040299719e-05,
+      "loss": 4.0076,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522923446126512e-05,
+      "loss": 4.0028,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52208485137546e-05,
+      "loss": 4.0014,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521246256624408e-05,
+      "loss": 4.0121,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520407661873356e-05,
+      "loss": 4.0032,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519570705002677e-05,
+      "loss": 4.0052,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518732110251625e-05,
+      "loss": 4.0076,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517893515500573e-05,
+      "loss": 4.0051,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517054920749521e-05,
+      "loss": 4.0065,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516216325998468e-05,
+      "loss": 4.0115,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153793691277896e-05,
+      "loss": 4.0139,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145407743767376e-05,
+      "loss": 4.0005,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137021796256856e-05,
+      "loss": 4.0124,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128635848746336e-05,
+      "loss": 4.0132,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512026628003955e-05,
+      "loss": 4.0022,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511188033252903e-05,
+      "loss": 4.0054,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510351076382224e-05,
+      "loss": 4.0081,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509512481631172e-05,
+      "loss": 4.0166,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50867388688012e-05,
+      "loss": 4.016,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507835292129068e-05,
+      "loss": 4.0078,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5069966973780154e-05,
+      "loss": 4.0079,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5061581026269634e-05,
+      "loss": 4.0112,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053195078759114e-05,
+      "loss": 4.0092,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044809131248594e-05,
+      "loss": 3.9976,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036423183738074e-05,
+      "loss": 4.0074,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028037236227554e-05,
+      "loss": 3.9939,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501965128871704e-05,
+      "loss": 4.0064,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501126534120652e-05,
+      "loss": 4.0057,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.047008991241455,
+      "eval_runtime": 298.4825,
+      "eval_samples_per_second": 1278.437,
+      "eval_steps_per_second": 39.952,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002879393696e-05,
+      "loss": 4.0164,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499449344618548e-05,
+      "loss": 3.9985,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498610749867496e-05,
+      "loss": 4.0043,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497772155116444e-05,
+      "loss": 4.0151,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496933560365392e-05,
+      "loss": 4.0115,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496094965614339e-05,
+      "loss": 4.0135,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495256370863287e-05,
+      "loss": 4.0087,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494417776112235e-05,
+      "loss": 3.9995,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493579181361183e-05,
+      "loss": 3.9941,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492740586610131e-05,
+      "loss": 4.0077,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491901991859079e-05,
+      "loss": 4.0056,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491063397108027e-05,
+      "loss": 4.0046,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490226440237349e-05,
+      "loss": 4.012,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489387845486297e-05,
+      "loss": 3.9931,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488549250735245e-05,
+      "loss": 4.0028,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487710655984193e-05,
+      "loss": 3.9839,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486872061233141e-05,
+      "loss": 4.0094,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486033466482089e-05,
+      "loss": 4.0027,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48519650961141e-05,
+      "loss": 4.0027,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484357914860358e-05,
+      "loss": 4.0047,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483519320109306e-05,
+      "loss": 4.013,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482680725358254e-05,
+      "loss": 4.0095,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481842130607202e-05,
+      "loss": 4.0034,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48100353585615e-05,
+      "loss": 4.0025,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480164941105098e-05,
+      "loss": 4.0154,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479326346354046e-05,
+      "loss": 4.0029,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478487751602994e-05,
+      "loss": 3.9961,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477650794732315e-05,
+      "loss": 4.0063,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476812199981263e-05,
+      "loss": 3.9938,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759736052302106e-05,
+      "loss": 4.0008,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751350104791586e-05,
+      "loss": 4.0068,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47429805360848e-05,
+      "loss": 3.9994,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473459458857428e-05,
+      "loss": 4.0058,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726208641063755e-05,
+      "loss": 4.013,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717822693553235e-05,
+      "loss": 3.9999,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470945312484645e-05,
+      "loss": 4.0082,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470106717733593e-05,
+      "loss": 4.0097,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469268122982541e-05,
+      "loss": 3.9916,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468429528231489e-05,
+      "loss": 3.9998,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467590933480437e-05,
+      "loss": 3.9901,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466752338729385e-05,
+      "loss": 3.9901,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465913743978333e-05,
+      "loss": 4.0006,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465075149227281e-05,
+      "loss": 4.008,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464236554476229e-05,
+      "loss": 3.9908,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463397959725177e-05,
+      "loss": 3.9989,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462561002854498e-05,
+      "loss": 4.0018,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461722408103446e-05,
+      "loss": 4.0009,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460885451232767e-05,
+      "loss": 3.9936,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460046856481715e-05,
+      "loss": 3.9826,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459208261730663e-05,
+      "loss": 4.0009,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458369666979611e-05,
+      "loss": 4.0135,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575310722285595e-05,
+      "loss": 4.011,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566924774775075e-05,
+      "loss": 3.9909,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558538827264555e-05,
+      "loss": 3.9967,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550152879754035e-05,
+      "loss": 4.0011,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.985,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533397363536724e-05,
+      "loss": 4.0168,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525011416026204e-05,
+      "loss": 3.9791,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516625468515684e-05,
+      "loss": 4.0179,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508239521005164e-05,
+      "loss": 4.0004,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4499853573494644e-05,
+      "loss": 3.9828,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449146762598412e-05,
+      "loss": 3.9991,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44830816784736e-05,
+      "loss": 3.9963,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447471210976681e-05,
+      "loss": 3.9775,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 4.0002,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 4.0047,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444957064603899e-05,
+      "loss": 3.9897,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444118469852847e-05,
+      "loss": 3.9995,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443281512982168e-05,
+      "loss": 3.9765,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442442918231116e-05,
+      "loss": 3.9913,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441604323480064e-05,
+      "loss": 3.99,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440765728729012e-05,
+      "loss": 3.9975,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439927133977959e-05,
+      "loss": 3.9912,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439088539226907e-05,
+      "loss": 3.997,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438249944475855e-05,
+      "loss": 4.0099,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437412987605177e-05,
+      "loss": 3.9968,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436574392854125e-05,
+      "loss": 3.9976,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435735798103073e-05,
+      "loss": 3.9947,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434897203352021e-05,
+      "loss": 3.9907,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434058608600969e-05,
+      "loss": 4.0015,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433220013849917e-05,
+      "loss": 3.9894,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432381419098865e-05,
+      "loss": 3.9969,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315444622281856e-05,
+      "loss": 4.001,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307058674771336e-05,
+      "loss": 3.9987,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298672727260816e-05,
+      "loss": 3.9746,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4290286779750296e-05,
+      "loss": 3.9975,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281917211043505e-05,
+      "loss": 3.9975,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4273531263532985e-05,
+      "loss": 3.9875,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265145316022465e-05,
+      "loss": 3.9989,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425677574731568e-05,
+      "loss": 3.9891,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424838979980516e-05,
+      "loss": 3.9867,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424000385229464e-05,
+      "loss": 3.9894,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423161790478412e-05,
+      "loss": 3.9914,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42232319572736e-05,
+      "loss": 4.0001,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421484600976308e-05,
+      "loss": 3.998,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420646006225256e-05,
+      "loss": 3.9994,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419807411474204e-05,
+      "loss": 3.9897,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418968816723152e-05,
+      "loss": 3.9855,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418131859852473e-05,
+      "loss": 3.9915,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417293265101421e-05,
+      "loss": 4.0001,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416454670350369e-05,
+      "loss": 3.9994,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415616075599317e-05,
+      "loss": 3.9949,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414777480848265e-05,
+      "loss": 3.9946,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139405239775865e-05,
+      "loss": 3.9849,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131019292265345e-05,
+      "loss": 3.9984,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122633344754825e-05,
+      "loss": 3.9896,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114247397244305e-05,
+      "loss": 3.9906,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410586144973378e-05,
+      "loss": 3.9861,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409747550222326e-05,
+      "loss": 3.9919,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408908955471274e-05,
+      "loss": 3.9899,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408070360720222e-05,
+      "loss": 3.986,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40723176596917e-05,
+      "loss": 3.9801,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406394809098491e-05,
+      "loss": 3.9929,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405556214347439e-05,
+      "loss": 3.9977,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047176195963874e-05,
+      "loss": 3.9887,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4038790248453354e-05,
+      "loss": 3.9975,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403042067974656e-05,
+      "loss": 4.0081,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9921,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401366516352925e-05,
+      "loss": 3.9899,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400527921601873e-05,
+      "loss": 3.9906,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399689326850821e-05,
+      "loss": 3.9947,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398852369980142e-05,
+      "loss": 3.9922,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39801377522909e-05,
+      "loss": 3.9901,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397175180478038e-05,
+      "loss": 3.9868,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396336585726986e-05,
+      "loss": 3.9952,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395497990975934e-05,
+      "loss": 3.9884,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946610341052556e-05,
+      "loss": 3.995,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938224393542036e-05,
+      "loss": 3.9921,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929838446031516e-05,
+      "loss": 3.9892,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921452498520996e-05,
+      "loss": 3.9872,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913066551010476e-05,
+      "loss": 4.0021,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904680603499956e-05,
+      "loss": 3.9966,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896311034793165e-05,
+      "loss": 3.9871,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3887925087282645e-05,
+      "loss": 3.9916,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879539139772125e-05,
+      "loss": 3.9997,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871153192261605e-05,
+      "loss": 3.9854,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862783623554814e-05,
+      "loss": 3.9952,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854397676044294e-05,
+      "loss": 3.9932,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384602810733751e-05,
+      "loss": 4.0011,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383764215982699e-05,
+      "loss": 3.9969,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382925621231647e-05,
+      "loss": 3.9973,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382087026480595e-05,
+      "loss": 3.9898,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381248431729543e-05,
+      "loss": 4.0007,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380409836978491e-05,
+      "loss": 3.99,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379572880107812e-05,
+      "loss": 3.9828,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37873428535676e-05,
+      "loss": 3.991,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377895690605708e-05,
+      "loss": 3.9782,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 3.993,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376218501103604e-05,
+      "loss": 3.9883,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.038602352142334,
+      "eval_runtime": 307.4727,
+      "eval_samples_per_second": 1241.057,
+      "eval_steps_per_second": 38.784,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375379906352552e-05,
+      "loss": 3.9927,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374542949481873e-05,
+      "loss": 3.9846,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3737043547308215e-05,
+      "loss": 3.989,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728657599797694e-05,
+      "loss": 4.0011,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720271652287174e-05,
+      "loss": 3.9944,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3711885704776654e-05,
+      "loss": 4.002,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703499757266134e-05,
+      "loss": 3.9895,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695113809755614e-05,
+      "loss": 3.9871,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368672786224509e-05,
+      "loss": 3.9768,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36783582935383e-05,
+      "loss": 3.9945,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366997234602778e-05,
+      "loss": 3.9934,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3661586398517256e-05,
+      "loss": 3.988,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365321682981047e-05,
+      "loss": 4.0013,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364483088229995e-05,
+      "loss": 3.974,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363644493478943e-05,
+      "loss": 3.9856,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362805898727891e-05,
+      "loss": 3.9777,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361967303976839e-05,
+      "loss": 3.9887,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361128709225787e-05,
+      "loss": 3.9904,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360290114474735e-05,
+      "loss": 3.9821,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359451519723683e-05,
+      "loss": 3.9945,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358612924972631e-05,
+      "loss": 3.9959,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357775968101952e-05,
+      "loss": 3.9943,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3569373733509e-05,
+      "loss": 3.9894,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356098778599848e-05,
+      "loss": 3.9873,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355260183848796e-05,
+      "loss": 4.0004,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354421589097744e-05,
+      "loss": 3.9861,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353584632227065e-05,
+      "loss": 3.9828,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352746037476014e-05,
+      "loss": 3.9893,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351907442724962e-05,
+      "loss": 3.9776,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35106884797391e-05,
+      "loss": 3.9843,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502318911032306e-05,
+      "loss": 3.9884,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3493932963521786e-05,
+      "loss": 3.9858,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485547016011266e-05,
+      "loss": 3.9937,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477161068500746e-05,
+      "loss": 3.9972,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468775120990226e-05,
+      "loss": 3.9858,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460405552283435e-05,
+      "loss": 3.993,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3452019604772915e-05,
+      "loss": 3.9957,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443633657262395e-05,
+      "loss": 3.9767,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435247709751875e-05,
+      "loss": 3.9833,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3426861762241354e-05,
+      "loss": 3.9718,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3418475814730834e-05,
+      "loss": 3.9795,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3410089867220314e-05,
+      "loss": 3.9837,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34017039197098e-05,
+      "loss": 3.9976,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3393317972199274e-05,
+      "loss": 3.9747,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338494840349249e-05,
+      "loss": 3.9828,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337656245598197e-05,
+      "loss": 3.9852,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336817650847144e-05,
+      "loss": 3.9843,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335979056096092e-05,
+      "loss": 3.981,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335143737105787e-05,
+      "loss": 3.9684,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334305142354735e-05,
+      "loss": 3.9873,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333466547603683e-05,
+      "loss": 3.9959,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3326295907330044e-05,
+      "loss": 3.9917,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3317909959819524e-05,
+      "loss": 3.9819,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309524012309004e-05,
+      "loss": 3.9789,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301138064798484e-05,
+      "loss": 3.9889,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292752117287964e-05,
+      "loss": 3.9696,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284366169777444e-05,
+      "loss": 4.0,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327598022226692e-05,
+      "loss": 3.9656,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32675942747564e-05,
+      "loss": 4.0026,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325920832724588e-05,
+      "loss": 3.9889,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325082237973536e-05,
+      "loss": 3.9661,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324243643222484e-05,
+      "loss": 3.983,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323405048471432e-05,
+      "loss": 3.9842,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32256645372038e-05,
+      "loss": 3.9658,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321727858969328e-05,
+      "loss": 3.9777,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320890902098649e-05,
+      "loss": 3.9955,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320052307347597e-05,
+      "loss": 3.9789,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319213712596545e-05,
+      "loss": 3.9813,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318375117845493e-05,
+      "loss": 3.9624,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317538160974814e-05,
+      "loss": 3.979,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316699566223762e-05,
+      "loss": 3.9736,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31586097147271e-05,
+      "loss": 3.9852,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315022376721658e-05,
+      "loss": 3.9778,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314185419850979e-05,
+      "loss": 3.9836,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313346825099927e-05,
+      "loss": 3.9917,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312508230348875e-05,
+      "loss": 3.9837,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311669635597823e-05,
+      "loss": 3.9809,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310831040846771e-05,
+      "loss": 3.9841,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309992446095719e-05,
+      "loss": 3.9794,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309153851344668e-05,
+      "loss": 3.985,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308315256593616e-05,
+      "loss": 3.9756,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3074782997229366e-05,
+      "loss": 3.9806,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3066397049718846e-05,
+      "loss": 3.9887,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058011102208326e-05,
+      "loss": 3.9833,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049625154697806e-05,
+      "loss": 3.9633,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304123920718728e-05,
+      "loss": 3.9826,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303285325967676e-05,
+      "loss": 3.9791,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302446731216624e-05,
+      "loss": 3.9788,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3016097743459455e-05,
+      "loss": 3.9787,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300771179594893e-05,
+      "loss": 3.9829,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2999325848438415e-05,
+      "loss": 3.9674,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2990939900927895e-05,
+      "loss": 3.9749,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298258671102484e-05,
+      "loss": 3.9742,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297420076351432e-05,
+      "loss": 3.9898,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29658148160038e-05,
+      "loss": 3.9809,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295742886849328e-05,
+      "loss": 3.9892,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294904292098275e-05,
+      "loss": 3.9715,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294065697347223e-05,
+      "loss": 3.9745,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293227102596171e-05,
+      "loss": 3.9768,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292388507845119e-05,
+      "loss": 3.9848,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291549913094067e-05,
+      "loss": 3.9841,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290712956223388e-05,
+      "loss": 3.9805,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289874361472337e-05,
+      "loss": 3.9816,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890374046016584e-05,
+      "loss": 3.9717,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288198809850606e-05,
+      "loss": 3.988,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287360215099554e-05,
+      "loss": 3.9728,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286521620348502e-05,
+      "loss": 3.9767,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28568302559745e-05,
+      "loss": 3.9743,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284844430846398e-05,
+      "loss": 3.9743,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284005836095346e-05,
+      "loss": 3.9755,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283167241344294e-05,
+      "loss": 3.9759,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282328646593242e-05,
+      "loss": 3.9633,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28149005184219e-05,
+      "loss": 3.9784,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280651457091138e-05,
+      "loss": 3.9824,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2798145002204586e-05,
+      "loss": 3.9731,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2789759054694066e-05,
+      "loss": 3.9879,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2781373107183546e-05,
+      "loss": 3.9947,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277298715967303e-05,
+      "loss": 3.9775,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276460121216251e-05,
+      "loss": 3.9768,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275621526465199e-05,
+      "loss": 3.9736,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2747829317141466e-05,
+      "loss": 3.9789,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273945974843468e-05,
+      "loss": 3.9764,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273107380092416e-05,
+      "loss": 3.9761,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272268785341364e-05,
+      "loss": 3.9734,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2714301905903115e-05,
+      "loss": 3.9823,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2705915958392595e-05,
+      "loss": 3.9769,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2697530010882075e-05,
+      "loss": 3.9778,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2689144063371555e-05,
+      "loss": 3.9771,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268077449466477e-05,
+      "loss": 3.9762,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2672404925957987e-05,
+      "loss": 3.9707,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2664018978447467e-05,
+      "loss": 3.9881,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265563303093694e-05,
+      "loss": 3.985,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264724708342642e-05,
+      "loss": 3.9772,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26388611359159e-05,
+      "loss": 3.9775,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263047518840538e-05,
+      "loss": 3.9835,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262208924089486e-05,
+      "loss": 3.9774,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261370329338434e-05,
+      "loss": 3.9774,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260531734587382e-05,
+      "loss": 3.9779,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25969313983633e-05,
+      "loss": 3.9937,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258856182965651e-05,
+      "loss": 3.9775,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258017588214599e-05,
+      "loss": 3.9858,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257178993463547e-05,
+      "loss": 3.9749,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563420365928684e-05,
+      "loss": 3.987,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2555034418418164e-05,
+      "loss": 3.9755,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2546648470907644e-05,
+      "loss": 3.9689,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2538262523397124e-05,
+      "loss": 3.9784,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2529876575886604e-05,
+      "loss": 3.9647,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2521490628376084e-05,
+      "loss": 3.9789,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2513104680865564e-05,
+      "loss": 3.9732,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.031342029571533,
+      "eval_runtime": 304.4621,
+      "eval_samples_per_second": 1253.328,
+      "eval_steps_per_second": 39.167,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504718733355044e-05,
+      "loss": 3.9741,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2496332785844524e-05,
+      "loss": 3.9763,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2487946838334004e-05,
+      "loss": 3.9729,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247956089082348e-05,
+      "loss": 3.9879,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247117494331296e-05,
+      "loss": 3.9784,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246278899580244e-05,
+      "loss": 3.9892,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2454403048291924e-05,
+      "loss": 3.9746,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2446017100781404e-05,
+      "loss": 3.9739,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2437631153270884e-05,
+      "loss": 3.9678,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2429245205760364e-05,
+      "loss": 3.9779,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2420859258249843e-05,
+      "loss": 3.9789,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2412473310739323e-05,
+      "loss": 3.9767,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240410374203253e-05,
+      "loss": 3.9893,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239571779452201e-05,
+      "loss": 3.9649,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238733184701149e-05,
+      "loss": 3.9673,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237894589950097e-05,
+      "loss": 3.9648,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237055995199045e-05,
+      "loss": 3.9732,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236217400447993e-05,
+      "loss": 3.9776,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235378805696941e-05,
+      "loss": 3.9733,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234541848826262e-05,
+      "loss": 3.9746,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23370325407521e-05,
+      "loss": 3.9848,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232864659324159e-05,
+      "loss": 3.9777,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232026064573107e-05,
+      "loss": 3.9802,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231189107702428e-05,
+      "loss": 3.9713,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230350512951376e-05,
+      "loss": 3.99,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229511918200324e-05,
+      "loss": 3.9731,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228673323449272e-05,
+      "loss": 3.9718,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227834728698219e-05,
+      "loss": 3.9715,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2269977718275406e-05,
+      "loss": 3.9668,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2261591770764886e-05,
+      "loss": 3.9723,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2253205823254366e-05,
+      "loss": 3.9751,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224481987574384e-05,
+      "loss": 3.9722,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2236433928233326e-05,
+      "loss": 3.9801,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2228047980722806e-05,
+      "loss": 3.9844,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2219662033212286e-05,
+      "loss": 3.9727,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2211292464505495e-05,
+      "loss": 3.9805,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2202906516994975e-05,
+      "loss": 3.9776,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2194520569484455e-05,
+      "loss": 3.9656,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2186134621973935e-05,
+      "loss": 3.9703,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2177748674463415e-05,
+      "loss": 3.9577,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2169362726952895e-05,
+      "loss": 3.9695,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2160976779442375e-05,
+      "loss": 3.9699,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2152590831931855e-05,
+      "loss": 3.9823,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2144204884421335e-05,
+      "loss": 3.9621,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2135835315714544e-05,
+      "loss": 3.9695,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2127449368204024e-05,
+      "loss": 3.9741,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211906342069351e-05,
+      "loss": 3.9696,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211067747318299e-05,
+      "loss": 3.9722,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210229152567247e-05,
+      "loss": 3.951,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209390557816195e-05,
+      "loss": 3.9712,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208551963065143e-05,
+      "loss": 3.985,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20771336831409e-05,
+      "loss": 3.9802,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206876411443412e-05,
+      "loss": 3.9675,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20603781669236e-05,
+      "loss": 3.9637,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205199221941308e-05,
+      "loss": 3.9764,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204360627190255e-05,
+      "loss": 3.96,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203522032439203e-05,
+      "loss": 3.9839,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202683437688151e-05,
+      "loss": 3.9518,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201844842937099e-05,
+      "loss": 3.9918,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201006248186048e-05,
+      "loss": 3.9759,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200169291315369e-05,
+      "loss": 3.9532,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199330696564317e-05,
+      "loss": 3.9654,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198492101813265e-05,
+      "loss": 3.9775,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197653507062213e-05,
+      "loss": 3.9506,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196816550191534e-05,
+      "loss": 3.9606,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195977955440482e-05,
+      "loss": 3.9849,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19513936068943e-05,
+      "loss": 3.9681,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1943024038187506e-05,
+      "loss": 3.9662,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1934638090676986e-05,
+      "loss": 3.947,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1926252143166466e-05,
+      "loss": 3.9661,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1917866195655946e-05,
+      "loss": 3.9619,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1909480248145426e-05,
+      "loss": 3.9758,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190109430063491e-05,
+      "loss": 3.9609,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189270835312439e-05,
+      "loss": 3.974,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188432240561387e-05,
+      "loss": 3.9763,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187593645810335e-05,
+      "loss": 3.9718,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186755051059283e-05,
+      "loss": 3.9677,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185918094188604e-05,
+      "loss": 3.9684,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185079499437552e-05,
+      "loss": 3.97,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1842409046865e-05,
+      "loss": 3.9737,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183402309935448e-05,
+      "loss": 3.9635,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182565353064769e-05,
+      "loss": 3.9667,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18172839619409e-05,
+      "loss": 3.9716,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180889801443038e-05,
+      "loss": 3.971,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1800512066919866e-05,
+      "loss": 3.9514,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1792126119409346e-05,
+      "loss": 3.9687,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1783740171898826e-05,
+      "loss": 3.964,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1775354224388306e-05,
+      "loss": 3.9674,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1766968276877786e-05,
+      "loss": 3.9686,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758582329367266e-05,
+      "loss": 3.9694,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175019638185674e-05,
+      "loss": 3.9577,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174181043434622e-05,
+      "loss": 3.9581,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17334244868357e-05,
+      "loss": 3.9606,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172503853932518e-05,
+      "loss": 3.9749,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171665259181466e-05,
+      "loss": 3.9685,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170826664430414e-05,
+      "loss": 3.9737,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169988069679362e-05,
+      "loss": 3.9614,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16914947492831e-05,
+      "loss": 3.9647,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1683141559380044e-05,
+      "loss": 3.9604,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167477199067326e-05,
+      "loss": 3.9752,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166638604316274e-05,
+      "loss": 3.9661,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165800009565221e-05,
+      "loss": 3.9693,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164961414814169e-05,
+      "loss": 3.9682,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164122820063117e-05,
+      "loss": 3.9578,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163284225312065e-05,
+      "loss": 3.9779,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162445630561013e-05,
+      "loss": 3.9563,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161608673690334e-05,
+      "loss": 3.9641,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160770078939282e-05,
+      "loss": 3.9659,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15993148418823e-05,
+      "loss": 3.9572,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159092889437179e-05,
+      "loss": 3.9627,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158254294686127e-05,
+      "loss": 3.9627,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157415699935075e-05,
+      "loss": 3.952,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156577105184023e-05,
+      "loss": 3.9673,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155738510432971e-05,
+      "loss": 3.9655,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154901553562292e-05,
+      "loss": 3.9634,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15406295881124e-05,
+      "loss": 3.9753,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153224364060188e-05,
+      "loss": 3.9784,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152385769309136e-05,
+      "loss": 3.9686,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151547174558084e-05,
+      "loss": 3.9639,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150708579807032e-05,
+      "loss": 3.9588,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498716229363526e-05,
+      "loss": 3.9647,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490330281853006e-05,
+      "loss": 3.9598,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1481944334342486e-05,
+      "loss": 3.9678,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147355838683197e-05,
+      "loss": 3.9557,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146518881812518e-05,
+      "loss": 3.9716,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145680287061466e-05,
+      "loss": 3.9614,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144841692310414e-05,
+      "loss": 3.9664,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144003097559362e-05,
+      "loss": 3.9649,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143166140688683e-05,
+      "loss": 3.9611,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142327545937631e-05,
+      "loss": 3.9567,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141488951186579e-05,
+      "loss": 3.975,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140650356435527e-05,
+      "loss": 3.9725,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139813399564848e-05,
+      "loss": 3.9699,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138974804813796e-05,
+      "loss": 3.9601,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138136210062744e-05,
+      "loss": 3.9673,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137297615311692e-05,
+      "loss": 3.9676,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13645902056064e-05,
+      "loss": 3.966,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135620425809588e-05,
+      "loss": 3.9691,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134781831058536e-05,
+      "loss": 3.976,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1339448741878575e-05,
+      "loss": 3.9664,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133106279436805e-05,
+      "loss": 3.9748,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132267684685753e-05,
+      "loss": 3.9635,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131429089934701e-05,
+      "loss": 3.9713,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130590495183649e-05,
+      "loss": 3.9579,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12975353831297e-05,
+      "loss": 3.9603,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128914943561918e-05,
+      "loss": 3.9661,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128076348810866e-05,
+      "loss": 3.9508,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1272377540598144e-05,
+      "loss": 3.9653,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1263991593087624e-05,
+      "loss": 3.9651,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.025163650512695,
+      "eval_runtime": 279.9135,
+      "eval_samples_per_second": 1363.246,
+      "eval_steps_per_second": 42.602,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1255605645577104e-05,
+      "loss": 3.9635,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124723607687031e-05,
+      "loss": 3.9634,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123885012935979e-05,
+      "loss": 3.96,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123046418184927e-05,
+      "loss": 3.973,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122207823433875e-05,
+      "loss": 3.9672,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121369228682823e-05,
+      "loss": 3.9762,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120530633931771e-05,
+      "loss": 3.9601,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119692039180719e-05,
+      "loss": 3.9621,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118853444429667e-05,
+      "loss": 3.9536,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118016487558988e-05,
+      "loss": 3.9695,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117177892807936e-05,
+      "loss": 3.9652,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116339298056884e-05,
+      "loss": 3.9643,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115500703305833e-05,
+      "loss": 3.9712,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114663746435154e-05,
+      "loss": 3.9567,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113825151684102e-05,
+      "loss": 3.9592,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11298655693305e-05,
+      "loss": 3.954,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112147962181998e-05,
+      "loss": 3.9538,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111309367430946e-05,
+      "loss": 3.9637,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110470772679894e-05,
+      "loss": 3.9618,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109632177928841e-05,
+      "loss": 3.962,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108793583177789e-05,
+      "loss": 3.9752,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107954988426737e-05,
+      "loss": 3.9638,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107118031556058e-05,
+      "loss": 3.9661,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1062810746853796e-05,
+      "loss": 3.9632,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105442479934328e-05,
+      "loss": 3.9749,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104603885183276e-05,
+      "loss": 3.9585,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1037652904322235e-05,
+      "loss": 3.9622,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1029266956811715e-05,
+      "loss": 3.9557,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1020881009301195e-05,
+      "loss": 3.9593,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1012495061790675e-05,
+      "loss": 3.9598,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1004125493083884e-05,
+      "loss": 3.9601,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0995739545573364e-05,
+      "loss": 3.9651,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0987353598062844e-05,
+      "loss": 3.9685,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0978967650552324e-05,
+      "loss": 3.9654,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0970581703041804e-05,
+      "loss": 3.9624,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096221213433502e-05,
+      "loss": 3.9672,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09538261868245e-05,
+      "loss": 3.9681,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094544023931398e-05,
+      "loss": 3.9509,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093705429180346e-05,
+      "loss": 3.9626,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092866834429294e-05,
+      "loss": 3.9428,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092028239678242e-05,
+      "loss": 3.9567,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09118964492719e-05,
+      "loss": 3.9539,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090351050176138e-05,
+      "loss": 3.9705,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089512455425086e-05,
+      "loss": 3.9529,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088673860674034e-05,
+      "loss": 3.9542,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087836903803355e-05,
+      "loss": 3.9617,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086999946932676e-05,
+      "loss": 3.9622,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086161352181624e-05,
+      "loss": 3.9585,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085322757430572e-05,
+      "loss": 3.9409,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08448416267952e-05,
+      "loss": 3.9562,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0836455679284685e-05,
+      "loss": 3.9734,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0828069731774164e-05,
+      "loss": 3.9678,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0819683784263644e-05,
+      "loss": 3.9589,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0811297836753124e-05,
+      "loss": 3.9522,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08029118892426e-05,
+      "loss": 3.9669,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0794542320535813e-05,
+      "loss": 3.9449,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0786156373025293e-05,
+      "loss": 3.9706,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07777868043185e-05,
+      "loss": 3.9416,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076940085680798e-05,
+      "loss": 3.9741,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076101490929746e-05,
+      "loss": 3.9672,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075262896178694e-05,
+      "loss": 3.9441,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074424301427642e-05,
+      "loss": 3.9544,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07358570667659e-05,
+      "loss": 3.966,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072747111925538e-05,
+      "loss": 3.935,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071908517174486e-05,
+      "loss": 3.9504,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071071560303807e-05,
+      "loss": 3.9718,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070232965552755e-05,
+      "loss": 3.9571,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069394370801703e-05,
+      "loss": 3.953,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.068555776050651e-05,
+      "loss": 3.9407,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067717181299599e-05,
+      "loss": 3.9527,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066878586548547e-05,
+      "loss": 3.9461,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066039991797495e-05,
+      "loss": 3.9638,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065201397046443e-05,
+      "loss": 3.9467,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064362802295391e-05,
+      "loss": 3.9632,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063524207544339e-05,
+      "loss": 3.9642,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062685612793287e-05,
+      "loss": 3.9589,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061847018042235e-05,
+      "loss": 3.9552,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0610116990519296e-05,
+      "loss": 3.9579,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0601731043008776e-05,
+      "loss": 3.9614,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0593345095498256e-05,
+      "loss": 3.9597,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0584959147987736e-05,
+      "loss": 3.9533,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0576573200477216e-05,
+      "loss": 3.953,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0568187252966696e-05,
+      "loss": 3.9592,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0559801305456176e-05,
+      "loss": 3.9604,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0551415357945655e-05,
+      "loss": 3.9402,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0543045789238865e-05,
+      "loss": 3.9549,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0534659841728345e-05,
+      "loss": 3.9521,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052629027302156e-05,
+      "loss": 3.9574,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051790432551104e-05,
+      "loss": 3.9591,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509518378000517e-05,
+      "loss": 3.9545,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501132430489997e-05,
+      "loss": 3.9488,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492746482979477e-05,
+      "loss": 3.944,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.048436053546896e-05,
+      "loss": 3.9495,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0475974587958433e-05,
+      "loss": 3.9633,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0467588640447913e-05,
+      "loss": 3.9574,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045921907174113e-05,
+      "loss": 3.9641,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0450833124230606e-05,
+      "loss": 3.9513,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044246355552382e-05,
+      "loss": 3.9507,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043409398681703e-05,
+      "loss": 3.9477,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042570803930651e-05,
+      "loss": 3.9638,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041732209179599e-05,
+      "loss": 3.9556,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040893614428547e-05,
+      "loss": 3.9603,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040055019677495e-05,
+      "loss": 3.9529,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.039216424926443e-05,
+      "loss": 3.9447,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383794680557643e-05,
+      "loss": 3.967,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0375408733047123e-05,
+      "loss": 3.9482,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0367022785536603e-05,
+      "loss": 3.9503,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0358636838026076e-05,
+      "loss": 3.9535,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035025089051556e-05,
+      "loss": 3.9532,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.034186494300504e-05,
+      "loss": 3.9479,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.033347899549452e-05,
+      "loss": 3.9493,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325093047984e-05,
+      "loss": 3.9432,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.031670710047348e-05,
+      "loss": 3.9536,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.030832115296296e-05,
+      "loss": 3.9529,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.029993520545244e-05,
+      "loss": 3.9546,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.029156563674565e-05,
+      "loss": 3.959,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.028317968923513e-05,
+      "loss": 3.9702,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.027479374172461e-05,
+      "loss": 3.9556,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.026640779421409e-05,
+      "loss": 3.9537,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.025802184670357e-05,
+      "loss": 3.9459,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.024963589919305e-05,
+      "loss": 3.9521,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.024126633048626e-05,
+      "loss": 3.9515,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0232880382975744e-05,
+      "loss": 3.9561,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224494435465224e-05,
+      "loss": 3.9454,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216108487954704e-05,
+      "loss": 3.9577,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0207738919247913e-05,
+      "loss": 3.9521,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0199352971737393e-05,
+      "loss": 3.9552,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0190967024226873e-05,
+      "loss": 3.9491,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182597455520085e-05,
+      "loss": 3.9518,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174211508009565e-05,
+      "loss": 3.9455,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0165825560499045e-05,
+      "loss": 3.9601,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157439612988525e-05,
+      "loss": 3.9653,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0149053665478005e-05,
+      "loss": 3.957,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140667717967485e-05,
+      "loss": 3.9469,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132298149260698e-05,
+      "loss": 3.9532,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0123912201750177e-05,
+      "loss": 3.9558,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115526254239657e-05,
+      "loss": 3.9569,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107140306729137e-05,
+      "loss": 3.9576,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098754359218617e-05,
+      "loss": 3.9639,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009036841170809e-05,
+      "loss": 3.9511,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0081982464197574e-05,
+      "loss": 3.9655,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073596516687054e-05,
+      "loss": 3.9492,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.006522694798027e-05,
+      "loss": 3.96,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056841000469743e-05,
+      "loss": 3.951,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048455052959223e-05,
+      "loss": 3.9446,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0040069105448703e-05,
+      "loss": 3.9577,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0031683157938183e-05,
+      "loss": 3.9432,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0023313589231395e-05,
+      "loss": 3.9488,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0014927641720875e-05,
+      "loss": 3.9556,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0006541694210355e-05,
+      "loss": 3.9525,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.018805027008057,
+      "eval_runtime": 303.8612,
+      "eval_samples_per_second": 1255.807,
+      "eval_steps_per_second": 39.245,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998155746699835e-05,
+      "loss": 3.9499,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989769799189315e-05,
+      "loss": 3.9507,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981383851678795e-05,
+      "loss": 3.9593,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9972997904168275e-05,
+      "loss": 3.9561,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964611956657755e-05,
+      "loss": 3.967,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956226009147238e-05,
+      "loss": 3.9507,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947840061636718e-05,
+      "loss": 3.9469,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939454114126198e-05,
+      "loss": 3.945,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931068166615678e-05,
+      "loss": 3.9555,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.992268221910515e-05,
+      "loss": 3.9524,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.991429627159463e-05,
+      "loss": 3.953,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9905910324084115e-05,
+      "loss": 3.9546,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989754075537733e-05,
+      "loss": 3.9466,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889154807866804e-05,
+      "loss": 3.9517,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880768860356284e-05,
+      "loss": 3.9422,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872382912845764e-05,
+      "loss": 3.9414,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9863996965335244e-05,
+      "loss": 3.9527,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855611017824723e-05,
+      "loss": 3.9501,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847241449117936e-05,
+      "loss": 3.9503,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9838855501607416e-05,
+      "loss": 3.9656,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830469554096896e-05,
+      "loss": 3.95,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822083606586376e-05,
+      "loss": 3.9555,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813697659075856e-05,
+      "loss": 3.9497,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805311711565336e-05,
+      "loss": 3.9637,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9796925764054816e-05,
+      "loss": 3.9495,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.97885398165443e-05,
+      "loss": 3.9515,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780170247837508e-05,
+      "loss": 3.944,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771784300326988e-05,
+      "loss": 3.9492,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763398352816468e-05,
+      "loss": 3.9443,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755012405305948e-05,
+      "loss": 3.9517,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974664283659916e-05,
+      "loss": 3.9542,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973825688908864e-05,
+      "loss": 3.9527,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972987094157812e-05,
+      "loss": 3.956,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97214849940676e-05,
+      "loss": 3.9529,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.971311542536081e-05,
+      "loss": 3.9499,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970472947785029e-05,
+      "loss": 3.9585,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969634353033977e-05,
+      "loss": 3.9428,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9687957582829253e-05,
+      "loss": 3.9477,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679571635318733e-05,
+      "loss": 3.9349,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9671185687808213e-05,
+      "loss": 3.9451,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662799740297693e-05,
+      "loss": 3.9439,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96544301715909e-05,
+      "loss": 3.9556,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.964604422408038e-05,
+      "loss": 3.9439,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963765827656986e-05,
+      "loss": 3.9438,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9629272329059345e-05,
+      "loss": 3.9528,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9620886381548818e-05,
+      "loss": 3.9497,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9612516812842034e-05,
+      "loss": 3.9429,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604130865331514e-05,
+      "loss": 3.9342,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959574491782099e-05,
+      "loss": 3.9393,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958735897031047e-05,
+      "loss": 3.9607,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957897302279995e-05,
+      "loss": 3.9589,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957058707528943e-05,
+      "loss": 3.9438,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.956220112777891e-05,
+      "loss": 3.9441,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.955381518026839e-05,
+      "loss": 3.9557,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.954542923275787e-05,
+      "loss": 3.9365,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537059664051083e-05,
+      "loss": 3.954,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528673716540563e-05,
+      "loss": 3.9328,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520287769030043e-05,
+      "loss": 3.965,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9511901821519523e-05,
+      "loss": 3.9531,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503532252812732e-05,
+      "loss": 3.9376,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949514630530221e-05,
+      "loss": 3.9379,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.948676035779169e-05,
+      "loss": 3.9579,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947837441028117e-05,
+      "loss": 3.9238,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9469988462770655e-05,
+      "loss": 3.9405,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461618894063864e-05,
+      "loss": 3.9563,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453232946553344e-05,
+      "loss": 3.9466,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9444846999042824e-05,
+      "loss": 3.9422,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436461051532304e-05,
+      "loss": 3.9303,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9428091482825516e-05,
+      "loss": 3.9378,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419721914118725e-05,
+      "loss": 3.9344,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411335966608205e-05,
+      "loss": 3.9555,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9402950019097685e-05,
+      "loss": 3.9362,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394564071587165e-05,
+      "loss": 3.9505,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386178124076645e-05,
+      "loss": 3.9487,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377792176566125e-05,
+      "loss": 3.9533,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936940622905561e-05,
+      "loss": 3.9402,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936102028154509e-05,
+      "loss": 3.9479,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935263433403457e-05,
+      "loss": 3.9489,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.934424838652405e-05,
+      "loss": 3.9487,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.933586243901353e-05,
+      "loss": 3.9414,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327476491503e-05,
+      "loss": 3.941,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319106922796217e-05,
+      "loss": 3.9458,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93107209752857e-05,
+      "loss": 3.9468,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302335027775174e-05,
+      "loss": 3.9342,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9293949080264654e-05,
+      "loss": 3.9398,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.928557951155787e-05,
+      "loss": 3.9427,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927719356404735e-05,
+      "loss": 3.9486,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9268807616536826e-05,
+      "loss": 3.9438,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260421669026306e-05,
+      "loss": 3.9438,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252052100319522e-05,
+      "loss": 3.936,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243666152809002e-05,
+      "loss": 3.9326,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235280205298475e-05,
+      "loss": 3.9383,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9226894257787955e-05,
+      "loss": 3.947,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921850831027744e-05,
+      "loss": 3.9447,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921012236276692e-05,
+      "loss": 3.9544,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.92017364152564e-05,
+      "loss": 3.9489,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.919335046774588e-05,
+      "loss": 3.9339,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9184980899039088e-05,
+      "loss": 3.9385,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91766113303323e-05,
+      "loss": 3.9481,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916822538282178e-05,
+      "loss": 3.9477,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915983943531126e-05,
+      "loss": 3.9504,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915145348780074e-05,
+      "loss": 3.9394,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.914306754029022e-05,
+      "loss": 3.9358,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91346815927797e-05,
+      "loss": 3.9556,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.912631202407291e-05,
+      "loss": 3.9345,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9117926076562392e-05,
+      "loss": 3.9414,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109540129051872e-05,
+      "loss": 3.9402,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101154181541352e-05,
+      "loss": 3.9458,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.909278461283456e-05,
+      "loss": 3.9334,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908439866532404e-05,
+      "loss": 3.9379,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.907601271781352e-05,
+      "loss": 3.9313,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067626770303e-05,
+      "loss": 3.9438,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059240822792484e-05,
+      "loss": 3.9401,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050854875281964e-05,
+      "loss": 3.9379,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9042485306575174e-05,
+      "loss": 3.9522,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034099359064653e-05,
+      "loss": 3.9589,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025713411554133e-05,
+      "loss": 3.9488,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017327464043613e-05,
+      "loss": 3.9387,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9008941516533093e-05,
+      "loss": 3.9372,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000555569022573e-05,
+      "loss": 3.9387,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992169621512057e-05,
+      "loss": 3.944,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8983783674001537e-05,
+      "loss": 3.9445,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8975414105294746e-05,
+      "loss": 3.9344,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967044536587955e-05,
+      "loss": 3.9456,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958658589077438e-05,
+      "loss": 3.9433,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950272641566918e-05,
+      "loss": 3.9447,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8941886694056398e-05,
+      "loss": 3.9397,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933500746545878e-05,
+      "loss": 3.9408,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925131177839087e-05,
+      "loss": 3.9371,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916745230328567e-05,
+      "loss": 3.9505,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908359282818047e-05,
+      "loss": 3.9527,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8899973335307527e-05,
+      "loss": 3.9507,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889158738779701e-05,
+      "loss": 3.9322,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888321781909022e-05,
+      "loss": 3.9471,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88748318715797e-05,
+      "loss": 3.9462,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886644592406918e-05,
+      "loss": 3.9462,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885805997655866e-05,
+      "loss": 3.94,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849674029048136e-05,
+      "loss": 3.9534,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841304460341352e-05,
+      "loss": 3.9402,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883291851283083e-05,
+      "loss": 3.9584,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8824532565320305e-05,
+      "loss": 3.9372,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8816146617809785e-05,
+      "loss": 3.9512,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807760670299265e-05,
+      "loss": 3.9411,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879939110159248e-05,
+      "loss": 3.9374,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791005154081957e-05,
+      "loss": 3.944,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782619206571437e-05,
+      "loss": 3.9317,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774233259060917e-05,
+      "loss": 3.9396,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765847311550397e-05,
+      "loss": 3.9426,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757461364039877e-05,
+      "loss": 3.9385,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.013398170471191,
+      "eval_runtime": 304.9301,
+      "eval_samples_per_second": 1251.405,
+      "eval_steps_per_second": 39.107,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8749075416529357e-05,
+      "loss": 3.9403,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.874068946901884e-05,
+      "loss": 3.9395,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.873230352150832e-05,
+      "loss": 3.9502,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.87239175739978e-05,
+      "loss": 3.9459,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.871553162648728e-05,
+      "loss": 3.9564,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.870714567897676e-05,
+      "loss": 3.939,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.869875973146624e-05,
+      "loss": 3.9393,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.869037378395572e-05,
+      "loss": 3.9339,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8681987836445197e-05,
+      "loss": 3.9463,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8673601888934677e-05,
+      "loss": 3.9414,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8665215941424157e-05,
+      "loss": 3.941,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8656829993913636e-05,
+      "loss": 3.9472,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8648460425206846e-05,
+      "loss": 3.9325,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8640074477696326e-05,
+      "loss": 3.9453,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.863168853018581e-05,
+      "loss": 3.9309,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.862330258267529e-05,
+      "loss": 3.9335,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861491663516477e-05,
+      "loss": 3.9387,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.860653068765425e-05,
+      "loss": 3.9368,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.859814474014373e-05,
+      "loss": 3.9387,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858975879263321e-05,
+      "loss": 3.9609,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858137284512269e-05,
+      "loss": 3.9371,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85730032764159e-05,
+      "loss": 3.9428,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.856461732890538e-05,
+      "loss": 3.9417,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.855623138139486e-05,
+      "loss": 3.9492,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.854784543388434e-05,
+      "loss": 3.9452,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853947586517755e-05,
+      "loss": 3.9445,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853108991766703e-05,
+      "loss": 3.9303,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.852270397015651e-05,
+      "loss": 3.9395,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.851431802264599e-05,
+      "loss": 3.931,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8505932075135473e-05,
+      "loss": 3.9442,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8497546127624953e-05,
+      "loss": 3.9383,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8489160180114433e-05,
+      "loss": 3.9424,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8480774232603913e-05,
+      "loss": 3.9507,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8472404663897122e-05,
+      "loss": 3.9383,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8464035095190335e-05,
+      "loss": 3.9457,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8455649147679815e-05,
+      "loss": 3.9435,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8447263200169295e-05,
+      "loss": 3.9357,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8438877252658775e-05,
+      "loss": 3.9366,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8430491305148255e-05,
+      "loss": 3.9253,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8422105357637735e-05,
+      "loss": 3.9329,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.841371941012721e-05,
+      "loss": 3.9363,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.840533346261669e-05,
+      "loss": 3.9452,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.839694751510617e-05,
+      "loss": 3.9337,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838856156759565e-05,
+      "loss": 3.933,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838017562008513e-05,
+      "loss": 3.9425,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.837178967257461e-05,
+      "loss": 3.9347,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8363436482671556e-05,
+      "loss": 3.9376,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8355050535161032e-05,
+      "loss": 3.9261,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8346680966454248e-05,
+      "loss": 3.9258,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8338295018943728e-05,
+      "loss": 3.9515,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8329909071433208e-05,
+      "loss": 3.9484,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.832152312392268e-05,
+      "loss": 3.9316,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8313137176412165e-05,
+      "loss": 3.9379,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8304751228901645e-05,
+      "loss": 3.9452,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8296365281391125e-05,
+      "loss": 3.9259,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8287979333880605e-05,
+      "loss": 3.9457,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279593386370085e-05,
+      "loss": 3.9221,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8271207438859565e-05,
+      "loss": 3.9513,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8262821491349044e-05,
+      "loss": 3.9443,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254435543838524e-05,
+      "loss": 3.9296,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8246065975131737e-05,
+      "loss": 3.9276,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8237680027621217e-05,
+      "loss": 3.9449,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8229294080110697e-05,
+      "loss": 3.9116,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8220908132600177e-05,
+      "loss": 3.9328,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8212522185089657e-05,
+      "loss": 3.946,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8204136237579137e-05,
+      "loss": 3.9392,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.819576666887235e-05,
+      "loss": 3.9297,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.818738072136183e-05,
+      "loss": 3.9201,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817899477385131e-05,
+      "loss": 3.9299,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817060882634079e-05,
+      "loss": 3.9289,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.816222287883027e-05,
+      "loss": 3.9396,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8153836931319742e-05,
+      "loss": 3.9293,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8145467362612958e-05,
+      "loss": 3.9401,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.813708141510244e-05,
+      "loss": 3.9419,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812869546759192e-05,
+      "loss": 3.9438,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8120309520081395e-05,
+      "loss": 3.9297,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8111923572570874e-05,
+      "loss": 3.9381,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.810355400386409e-05,
+      "loss": 3.9396,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.809516805635357e-05,
+      "loss": 3.938,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8086782108843047e-05,
+      "loss": 3.9337,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8078412540136263e-05,
+      "loss": 3.931,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8070026592625743e-05,
+      "loss": 3.94,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8061640645115216e-05,
+      "loss": 3.9368,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8053254697604696e-05,
+      "loss": 3.921,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804486875009418e-05,
+      "loss": 3.9295,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803648280258366e-05,
+      "loss": 3.9306,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8028113233876868e-05,
+      "loss": 3.9408,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8019727286366348e-05,
+      "loss": 3.9366,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8011341338855828e-05,
+      "loss": 3.9306,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8002955391345308e-05,
+      "loss": 3.9272,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7994569443834788e-05,
+      "loss": 3.9236,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798618349632427e-05,
+      "loss": 3.9271,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.797779754881375e-05,
+      "loss": 3.9381,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796941160130323e-05,
+      "loss": 3.9398,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796104203259644e-05,
+      "loss": 3.9428,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.795265608508592e-05,
+      "loss": 3.9383,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.79442701375754e-05,
+      "loss": 3.928,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.793588419006488e-05,
+      "loss": 3.9299,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7927514621358093e-05,
+      "loss": 3.9324,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919128673847573e-05,
+      "loss": 3.9407,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7910742726337053e-05,
+      "loss": 3.9384,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7902356778826533e-05,
+      "loss": 3.9327,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7893987210119742e-05,
+      "loss": 3.9246,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.788560126260922e-05,
+      "loss": 3.9437,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7877215315098705e-05,
+      "loss": 3.9233,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7868829367588185e-05,
+      "loss": 3.9334,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7860443420077665e-05,
+      "loss": 3.9332,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7852057472567145e-05,
+      "loss": 3.9342,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7843671525056625e-05,
+      "loss": 3.9228,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7835285577546105e-05,
+      "loss": 3.9326,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7826916008839314e-05,
+      "loss": 3.9208,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7818530061328797e-05,
+      "loss": 3.9318,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7810144113818277e-05,
+      "loss": 3.9322,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7801774545111486e-05,
+      "loss": 3.9258,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7793388597600966e-05,
+      "loss": 3.947,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7785002650090446e-05,
+      "loss": 3.9487,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7776616702579926e-05,
+      "loss": 3.9388,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7768230755069406e-05,
+      "loss": 3.9304,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7759844807558883e-05,
+      "loss": 3.9236,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7751458860048363e-05,
+      "loss": 3.9303,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7743072912537843e-05,
+      "loss": 3.9293,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7734686965027323e-05,
+      "loss": 3.941,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7726301017516803e-05,
+      "loss": 3.9205,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7717931448810015e-05,
+      "loss": 3.9325,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7709545501299495e-05,
+      "loss": 3.9354,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7701159553788975e-05,
+      "loss": 3.9384,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7692773606278455e-05,
+      "loss": 3.9241,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7684387658767935e-05,
+      "loss": 3.9317,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7676001711257415e-05,
+      "loss": 3.93,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7667632142550627e-05,
+      "loss": 3.9349,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7659246195040107e-05,
+      "loss": 3.9426,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7650876626333316e-05,
+      "loss": 3.9406,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7642490678822796e-05,
+      "loss": 3.9246,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7634104731312276e-05,
+      "loss": 3.936,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7625718783801756e-05,
+      "loss": 3.9381,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7617332836291236e-05,
+      "loss": 3.9354,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760894688878072e-05,
+      "loss": 3.9331,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76005609412702e-05,
+      "loss": 3.9405,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.759217499375968e-05,
+      "loss": 3.9333,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758380542505289e-05,
+      "loss": 3.9486,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.757541947754237e-05,
+      "loss": 3.9288,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.756703353003185e-05,
+      "loss": 3.9401,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755864758252133e-05,
+      "loss": 3.9327,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755027801381454e-05,
+      "loss": 3.9291,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.754189206630402e-05,
+      "loss": 3.9337,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.75335061187935e-05,
+      "loss": 3.928,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.752512017128298e-05,
+      "loss": 3.9295,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.751673422377246e-05,
+      "loss": 3.9315,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.750834827626194e-05,
+      "loss": 3.929,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.008862495422363,
+      "eval_runtime": 296.4449,
+      "eval_samples_per_second": 1287.224,
+      "eval_steps_per_second": 40.227,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7499962328751417e-05,
+      "loss": 3.9343,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7491592760044633e-05,
+      "loss": 3.9283,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7483223191337842e-05,
+      "loss": 3.9412,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7474837243827322e-05,
+      "loss": 3.9333,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7466451296316802e-05,
+      "loss": 3.9491,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7458065348806282e-05,
+      "loss": 3.9273,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7449679401295762e-05,
+      "loss": 3.9321,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.744129345378524e-05,
+      "loss": 3.9208,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.743290750627472e-05,
+      "loss": 3.9383,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.74245215587642e-05,
+      "loss": 3.933,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741613561125368e-05,
+      "loss": 3.9286,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.740774966374316e-05,
+      "loss": 3.9439,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739938009503637e-05,
+      "loss": 3.9216,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739099414752585e-05,
+      "loss": 3.9368,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.738260820001533e-05,
+      "loss": 3.9207,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.737422225250481e-05,
+      "loss": 3.9243,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.736583630499429e-05,
+      "loss": 3.9229,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7357483115091236e-05,
+      "loss": 3.9333,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7349097167580712e-05,
+      "loss": 3.9276,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7340711220070192e-05,
+      "loss": 3.9502,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7332325272559672e-05,
+      "loss": 3.9293,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7323939325049152e-05,
+      "loss": 3.9331,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7315553377538632e-05,
+      "loss": 3.9318,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307167430028112e-05,
+      "loss": 3.9411,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7298781482517592e-05,
+      "loss": 3.9302,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290395535007075e-05,
+      "loss": 3.9359,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7282009587496555e-05,
+      "loss": 3.9249,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7273623639986035e-05,
+      "loss": 3.9292,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7265237692475515e-05,
+      "loss": 3.919,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7256868123768724e-05,
+      "loss": 3.9365,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7248482176258204e-05,
+      "loss": 3.9329,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7240112607551417e-05,
+      "loss": 3.9322,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7231726660040897e-05,
+      "loss": 3.9417,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7223340712530377e-05,
+      "loss": 3.9298,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7214971143823586e-05,
+      "loss": 3.9302,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7206585196313066e-05,
+      "loss": 3.9365,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7198199248802546e-05,
+      "loss": 3.9284,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.718981330129203e-05,
+      "loss": 3.926,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.718142735378151e-05,
+      "loss": 3.9154,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.717304140627099e-05,
+      "loss": 3.9225,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7164671837564198e-05,
+      "loss": 3.926,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7156285890053678e-05,
+      "loss": 3.9341,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7147899942543158e-05,
+      "loss": 3.9285,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7139513995032638e-05,
+      "loss": 3.9215,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.713112804752212e-05,
+      "loss": 3.9371,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.71227421000116e-05,
+      "loss": 3.9283,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114356152501074e-05,
+      "loss": 3.9264,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7105970204990554e-05,
+      "loss": 3.9209,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097584257480034e-05,
+      "loss": 3.9145,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.708921468877325e-05,
+      "loss": 3.9447,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7080828741262727e-05,
+      "loss": 3.9387,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7072442793752207e-05,
+      "loss": 3.9256,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7064056846241687e-05,
+      "loss": 3.9259,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7055670898731167e-05,
+      "loss": 3.9367,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7047301330024376e-05,
+      "loss": 3.9159,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703891538251386e-05,
+      "loss": 3.9359,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703052943500334e-05,
+      "loss": 3.9133,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.702214348749282e-05,
+      "loss": 3.9359,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.70137575399823e-05,
+      "loss": 3.9402,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700537159247178e-05,
+      "loss": 3.9234,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.699698564496126e-05,
+      "loss": 3.9161,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698859969745074e-05,
+      "loss": 3.9368,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698023012874395e-05,
+      "loss": 3.9072,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.697184418123343e-05,
+      "loss": 3.9251,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.696345823372291e-05,
+      "loss": 3.9327,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.695507228621239e-05,
+      "loss": 3.933,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.694668633870187e-05,
+      "loss": 3.9253,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.693830039119135e-05,
+      "loss": 3.9123,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692991444368083e-05,
+      "loss": 3.9168,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6921544874974043e-05,
+      "loss": 3.9172,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6913158927463523e-05,
+      "loss": 3.9314,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6904772979953003e-05,
+      "loss": 3.9211,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6896387032442483e-05,
+      "loss": 3.9317,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6888001084931963e-05,
+      "loss": 3.9329,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6879631516225172e-05,
+      "loss": 3.9353,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6871245568714652e-05,
+      "loss": 3.9186,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6862859621204132e-05,
+      "loss": 3.929,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6854473673693616e-05,
+      "loss": 3.9332,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.684608772618309e-05,
+      "loss": 3.9235,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6837718157476305e-05,
+      "loss": 3.9279,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6829348588769514e-05,
+      "loss": 3.9246,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6820962641258994e-05,
+      "loss": 3.9241,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6812576693748477e-05,
+      "loss": 3.931,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6804190746237957e-05,
+      "loss": 3.9114,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6795804798727437e-05,
+      "loss": 3.9246,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6787435230020646e-05,
+      "loss": 3.9189,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6779049282510126e-05,
+      "loss": 3.9331,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6770663334999606e-05,
+      "loss": 3.9262,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6762277387489086e-05,
+      "loss": 3.9235,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6753891439978563e-05,
+      "loss": 3.9186,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6745505492468042e-05,
+      "loss": 3.9171,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6737119544957522e-05,
+      "loss": 3.914,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6728733597447002e-05,
+      "loss": 3.9256,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720347649936482e-05,
+      "loss": 3.9378,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6711978081229695e-05,
+      "loss": 3.9302,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6703592133719175e-05,
+      "loss": 3.9245,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6695206186208655e-05,
+      "loss": 3.9169,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6686836617501864e-05,
+      "loss": 3.9212,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6678450669991344e-05,
+      "loss": 3.9223,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6670064722480824e-05,
+      "loss": 3.9345,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6661678774970307e-05,
+      "loss": 3.9289,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6653292827459787e-05,
+      "loss": 3.9239,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6644906879949267e-05,
+      "loss": 3.9153,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6636520932438747e-05,
+      "loss": 3.9361,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6628134984928227e-05,
+      "loss": 3.9156,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.661978179502517e-05,
+      "loss": 3.925,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.661139584751465e-05,
+      "loss": 3.9194,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.660300990000413e-05,
+      "loss": 3.9262,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.659462395249361e-05,
+      "loss": 3.9162,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.658623800498309e-05,
+      "loss": 3.9257,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6577852057472568e-05,
+      "loss": 3.9142,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6569466109962048e-05,
+      "loss": 3.9184,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6561080162451528e-05,
+      "loss": 3.9264,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6552694214941008e-05,
+      "loss": 3.9161,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.654430826743049e-05,
+      "loss": 3.9389,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.65359386987237e-05,
+      "loss": 3.9377,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652756913001691e-05,
+      "loss": 3.9303,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651918318250639e-05,
+      "loss": 3.9215,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651079723499587e-05,
+      "loss": 3.9192,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502411287485353e-05,
+      "loss": 3.9213,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6494025339974833e-05,
+      "loss": 3.9214,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6485639392464313e-05,
+      "loss": 3.9323,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6477253444953793e-05,
+      "loss": 3.9141,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6468867497443273e-05,
+      "loss": 3.9227,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6460514307540214e-05,
+      "loss": 3.9284,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6452128360029694e-05,
+      "loss": 3.9289,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6443742412519174e-05,
+      "loss": 3.9185,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435356465008654e-05,
+      "loss": 3.9212,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6426970517498134e-05,
+      "loss": 3.924,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6418584569987614e-05,
+      "loss": 3.9256,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6410198622477094e-05,
+      "loss": 3.9296,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.640181267496657e-05,
+      "loss": 3.9339,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639342672745605e-05,
+      "loss": 3.9131,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6385057158749267e-05,
+      "loss": 3.9297,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6376671211238746e-05,
+      "loss": 3.9315,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636828526372822e-05,
+      "loss": 3.9247,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.63598993162177e-05,
+      "loss": 3.9234,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6351513368707183e-05,
+      "loss": 3.9355,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6343127421196663e-05,
+      "loss": 3.9238,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6334757852489872e-05,
+      "loss": 3.9397,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6326371904979352e-05,
+      "loss": 3.9202,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6318002336272568e-05,
+      "loss": 3.929,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6309616388762044e-05,
+      "loss": 3.9253,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6301230441251524e-05,
+      "loss": 3.9223,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6292844493741004e-05,
+      "loss": 3.9227,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284458546230484e-05,
+      "loss": 3.9217,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6276072598719964e-05,
+      "loss": 3.9177,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6267703030013173e-05,
+      "loss": 3.9218,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6259317082502653e-05,
+      "loss": 3.9225,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.005211353302002,
+      "eval_runtime": 295.3661,
+      "eval_samples_per_second": 1291.926,
+      "eval_steps_per_second": 40.374,
+      "step": 1450080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6250931134992137e-05,
+      "loss": 3.9261,
+      "step": 1450496
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6242545187481617e-05,
+      "loss": 3.9199,
+      "step": 1451008
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6234159239971097e-05,
+      "loss": 3.9324,
+      "step": 1451520
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6225773292460577e-05,
+      "loss": 3.9229,
+      "step": 1452032
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6217387344950056e-05,
+      "loss": 3.9432,
+      "step": 1452544
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6209001397439536e-05,
+      "loss": 3.9219,
+      "step": 1453056
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6200615449929016e-05,
+      "loss": 3.9231,
+      "step": 1453568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6192229502418496e-05,
+      "loss": 3.9105,
+      "step": 1454080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.618385993371171e-05,
+      "loss": 3.928,
+      "step": 1454592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.617547398620119e-05,
+      "loss": 3.9274,
+      "step": 1455104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.616708803869067e-05,
+      "loss": 3.9193,
+      "step": 1455616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.615870209118015e-05,
+      "loss": 3.9293,
+      "step": 1456128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6150332522473358e-05,
+      "loss": 3.9182,
+      "step": 1456640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6141946574962838e-05,
+      "loss": 3.9285,
+      "step": 1457152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6133560627452318e-05,
+      "loss": 3.9154,
+      "step": 1457664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.61251746799418e-05,
+      "loss": 3.9127,
+      "step": 1458176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.611678873243128e-05,
+      "loss": 3.9185,
+      "step": 1458688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6108402784920754e-05,
+      "loss": 3.9226,
+      "step": 1459200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6100016837410234e-05,
+      "loss": 3.9209,
+      "step": 1459712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6091630889899714e-05,
+      "loss": 3.9397,
+      "step": 1460224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6083244942389194e-05,
+      "loss": 3.9219,
+      "step": 1460736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6074875373682407e-05,
+      "loss": 3.9308,
+      "step": 1461248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6066489426171886e-05,
+      "loss": 3.9223,
+      "step": 1461760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6058103478661366e-05,
+      "loss": 3.9307,
+      "step": 1462272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6049733909954582e-05,
+      "loss": 3.9207,
+      "step": 1462784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6041347962444055e-05,
+      "loss": 3.927,
+      "step": 1463296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.603296201493354e-05,
+      "loss": 3.9187,
+      "step": 1463808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.602457606742302e-05,
+      "loss": 3.9165,
+      "step": 1464320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.60161901199125e-05,
+      "loss": 3.9153,
+      "step": 1464832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.600780417240198e-05,
+      "loss": 3.9243,
+      "step": 1465344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.599941822489146e-05,
+      "loss": 3.927,
+      "step": 1465856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5991048656184668e-05,
+      "loss": 3.9233,
+      "step": 1466368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5982662708674148e-05,
+      "loss": 3.9351,
+      "step": 1466880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.597427676116363e-05,
+      "loss": 3.9221,
+      "step": 1467392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.596590719245684e-05,
+      "loss": 3.9192,
+      "step": 1467904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.595752124494632e-05,
+      "loss": 3.9273,
+      "step": 1468416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.59491352974358e-05,
+      "loss": 3.9265,
+      "step": 1468928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.594074934992528e-05,
+      "loss": 3.9127,
+      "step": 1469440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.593236340241476e-05,
+      "loss": 3.9056,
+      "step": 1469952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.592397745490424e-05,
+      "loss": 3.911,
+      "step": 1470464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5915591507393723e-05,
+      "loss": 3.9245,
+      "step": 1470976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5907205559883203e-05,
+      "loss": 3.9246,
+      "step": 1471488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5898819612372683e-05,
+      "loss": 3.9199,
+      "step": 1472000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5890433664862163e-05,
+      "loss": 3.9105,
+      "step": 1472512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5882047717351643e-05,
+      "loss": 3.9273,
+      "step": 1473024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5873661769841116e-05,
+      "loss": 3.919,
+      "step": 1473536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5865292201134332e-05,
+      "loss": 3.9188,
+      "step": 1474048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5856906253623815e-05,
+      "loss": 3.9117,
+      "step": 1474560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5848536684917025e-05,
+      "loss": 3.8989,
+      "step": 1475072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5840150737406505e-05,
+      "loss": 3.9435,
+      "step": 1475584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5831764789895984e-05,
+      "loss": 3.927,
+      "step": 1476096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5823378842385464e-05,
+      "loss": 3.9193,
+      "step": 1476608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.581499289487494e-05,
+      "loss": 3.9205,
+      "step": 1477120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.580660694736442e-05,
+      "loss": 3.9241,
+      "step": 1477632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.57982209998539e-05,
+      "loss": 3.9115,
+      "step": 1478144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.578983505234338e-05,
+      "loss": 3.923,
+      "step": 1478656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.578144910483286e-05,
+      "loss": 3.9101,
+      "step": 1479168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.577306315732234e-05,
+      "loss": 3.9239,
+      "step": 1479680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.576467720981182e-05,
+      "loss": 3.9345,
+      "step": 1480192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.57562912623013e-05,
+      "loss": 3.9137,
+      "step": 1480704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.574790531479078e-05,
+      "loss": 3.9036,
+      "step": 1481216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5739535746083993e-05,
+      "loss": 3.9334,
+      "step": 1481728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5731149798573473e-05,
+      "loss": 3.8991,
+      "step": 1482240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5722763851062953e-05,
+      "loss": 3.9129,
+      "step": 1482752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5714377903552433e-05,
+      "loss": 3.9289,
+      "step": 1483264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5705991956041913e-05,
+      "loss": 3.9243,
+      "step": 1483776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5697606008531393e-05,
+      "loss": 3.9174,
+      "step": 1484288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5689236439824605e-05,
+      "loss": 3.9089,
+      "step": 1484800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5680850492314085e-05,
+      "loss": 3.9021,
+      "step": 1485312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5672480923607294e-05,
+      "loss": 3.9083,
+      "step": 1485824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5664094976096774e-05,
+      "loss": 3.9257,
+      "step": 1486336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5655709028586254e-05,
+      "loss": 3.9095,
+      "step": 1486848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5647323081075734e-05,
+      "loss": 3.9269,
+      "step": 1487360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5638937133565218e-05,
+      "loss": 3.9212,
+      "step": 1487872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5630551186054698e-05,
+      "loss": 3.9318,
+      "step": 1488384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5622165238544178e-05,
+      "loss": 3.9126,
+      "step": 1488896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5613779291033658e-05,
+      "loss": 3.9172,
+      "step": 1489408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.560539334352313e-05,
+      "loss": 3.9253,
+      "step": 1489920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5597023774816347e-05,
+      "loss": 3.9134,
+      "step": 1490432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5588637827305827e-05,
+      "loss": 3.9256,
+      "step": 1490944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5580251879795303e-05,
+      "loss": 3.9149,
+      "step": 1491456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5571865932284783e-05,
+      "loss": 3.9163,
+      "step": 1491968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5563479984774263e-05,
+      "loss": 3.922,
+      "step": 1492480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.555511041606748e-05,
+      "loss": 3.9043,
+      "step": 1492992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5546724468556955e-05,
+      "loss": 3.9153,
+      "step": 1493504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5538338521046435e-05,
+      "loss": 3.9146,
+      "step": 1494016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5529952573535915e-05,
+      "loss": 3.9214,
+      "step": 1494528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5521566626025395e-05,
+      "loss": 3.9156,
+      "step": 1495040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5513197057318604e-05,
+      "loss": 3.9163,
+      "step": 1495552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5504811109808084e-05,
+      "loss": 3.9082,
+      "step": 1496064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5496425162297564e-05,
+      "loss": 3.9093,
+      "step": 1496576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5488039214787048e-05,
+      "loss": 3.9093,
+      "step": 1497088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5479653267276528e-05,
+      "loss": 3.9154,
+      "step": 1497600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5471283698569737e-05,
+      "loss": 3.9309,
+      "step": 1498112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5462897751059217e-05,
+      "loss": 3.9258,
+      "step": 1498624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5454511803548697e-05,
+      "loss": 3.9195,
+      "step": 1499136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5446125856038177e-05,
+      "loss": 3.9072,
+      "step": 1499648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.543775628733139e-05,
+      "loss": 3.9142,
+      "step": 1500160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.542937033982087e-05,
+      "loss": 3.9167,
+      "step": 1500672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.542098439231035e-05,
+      "loss": 3.9229,
+      "step": 1501184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.541259844479983e-05,
+      "loss": 3.9211,
+      "step": 1501696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.540421249728931e-05,
+      "loss": 3.916,
+      "step": 1502208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.539582654977879e-05,
+      "loss": 3.9071,
+      "step": 1502720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.538744060226827e-05,
+      "loss": 3.931,
+      "step": 1503232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.537905465475775e-05,
+      "loss": 3.9041,
+      "step": 1503744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.537068508605096e-05,
+      "loss": 3.9201,
+      "step": 1504256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.536231551734417e-05,
+      "loss": 3.9107,
+      "step": 1504768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.535392956983365e-05,
+      "loss": 3.9184,
+      "step": 1505280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.534554362232313e-05,
+      "loss": 3.9094,
+      "step": 1505792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.533715767481261e-05,
+      "loss": 3.917,
+      "step": 1506304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5328771727302094e-05,
+      "loss": 3.9097,
+      "step": 1506816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5320385779791574e-05,
+      "loss": 3.9108,
+      "step": 1507328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5311999832281053e-05,
+      "loss": 3.9158,
+      "step": 1507840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5303613884770533e-05,
+      "loss": 3.9107,
+      "step": 1508352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5295244316063743e-05,
+      "loss": 3.9278,
+      "step": 1508864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5286858368553222e-05,
+      "loss": 3.9313,
+      "step": 1509376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5278488799846435e-05,
+      "loss": 3.924,
+      "step": 1509888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5270102852335915e-05,
+      "loss": 3.9122,
+      "step": 1510400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5261716904825395e-05,
+      "loss": 3.9124,
+      "step": 1510912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5253330957314875e-05,
+      "loss": 3.912,
+      "step": 1511424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5244945009804355e-05,
+      "loss": 3.913,
+      "step": 1511936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5236575441097564e-05,
+      "loss": 3.9263,
+      "step": 1512448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5228189493587047e-05,
+      "loss": 3.9012,
+      "step": 1512960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5219819924880256e-05,
+      "loss": 3.9194,
+      "step": 1513472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5211433977369736e-05,
+      "loss": 3.916,
+      "step": 1513984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5203048029859216e-05,
+      "loss": 3.923,
+      "step": 1514496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5194662082348696e-05,
+      "loss": 3.9104,
+      "step": 1515008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.518629251364191e-05,
+      "loss": 3.9161,
+      "step": 1515520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.517790656613139e-05,
+      "loss": 3.9111,
+      "step": 1516032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.516952061862087e-05,
+      "loss": 3.925,
+      "step": 1516544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.516113467111035e-05,
+      "loss": 3.9201,
+      "step": 1517056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.515274872359983e-05,
+      "loss": 3.9292,
+      "step": 1517568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.514436277608931e-05,
+      "loss": 3.9041,
+      "step": 1518080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.513597682857879e-05,
+      "loss": 3.9215,
+      "step": 1518592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5127590881068265e-05,
+      "loss": 3.9231,
+      "step": 1519104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5119204933557745e-05,
+      "loss": 3.9129,
+      "step": 1519616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5110818986047225e-05,
+      "loss": 3.9173,
+      "step": 1520128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5102433038536705e-05,
+      "loss": 3.9264,
+      "step": 1520640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5094047091026185e-05,
+      "loss": 3.9175,
+      "step": 1521152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5085677522319394e-05,
+      "loss": 3.9346,
+      "step": 1521664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.507730795361261e-05,
+      "loss": 3.91,
+      "step": 1522176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5068922006102086e-05,
+      "loss": 3.9222,
+      "step": 1522688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5060536058591566e-05,
+      "loss": 3.9124,
+      "step": 1523200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5052150111081046e-05,
+      "loss": 3.9179,
+      "step": 1523712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5043764163570526e-05,
+      "loss": 3.9156,
+      "step": 1524224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5035378216060006e-05,
+      "loss": 3.9119,
+      "step": 1524736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5026992268549486e-05,
+      "loss": 3.912,
+      "step": 1525248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.50186226998427e-05,
+      "loss": 3.9142,
+      "step": 1525760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.501023675233218e-05,
+      "loss": 3.9149,
+      "step": 1526272
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.001781463623047,
+      "eval_runtime": 293.4498,
+      "eval_samples_per_second": 1300.362,
+      "eval_steps_per_second": 40.637,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.500185080482166e-05,
+      "loss": 3.9173,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.499346485731114e-05,
+      "loss": 3.9097,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.498507890980062e-05,
+      "loss": 3.9252,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49766929622901e-05,
+      "loss": 3.9167,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.496830701477958e-05,
+      "loss": 3.9316,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4959921067269058e-05,
+      "loss": 3.9167,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951535119758538e-05,
+      "loss": 3.9153,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4943149172248018e-05,
+      "loss": 3.9035,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4934763224737498e-05,
+      "loss": 3.9179,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4926377277226978e-05,
+      "loss": 3.9184,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4917991329716458e-05,
+      "loss": 3.9168,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909605382205938e-05,
+      "loss": 3.9191,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.490123581349915e-05,
+      "loss": 3.9084,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.489284986598863e-05,
+      "loss": 3.923,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.488446391847811e-05,
+      "loss": 3.9063,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.487607797096759e-05,
+      "loss": 3.9043,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4867692023457067e-05,
+      "loss": 3.9098,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859306075946547e-05,
+      "loss": 3.916,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4850920128436027e-05,
+      "loss": 3.9115,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.484253418092551e-05,
+      "loss": 3.9305,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.483416461221872e-05,
+      "loss": 3.9143,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.48257786647082e-05,
+      "loss": 3.9247,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.481739271719768e-05,
+      "loss": 3.9126,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.480900676968716e-05,
+      "loss": 3.92,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.480062082217664e-05,
+      "loss": 3.9135,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.479223487466612e-05,
+      "loss": 3.922,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.47838489271556e-05,
+      "loss": 3.9099,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.477546297964508e-05,
+      "loss": 3.9115,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.476709341093829e-05,
+      "loss": 3.9069,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.475870746342777e-05,
+      "loss": 3.9139,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750321515917248e-05,
+      "loss": 3.9207,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4741951947210464e-05,
+      "loss": 3.9164,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733565999699944e-05,
+      "loss": 3.925,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.472518005218942e-05,
+      "loss": 3.9115,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.47167941046789e-05,
+      "loss": 3.9144,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4708424535972113e-05,
+      "loss": 3.9243,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4700038588461593e-05,
+      "loss": 3.9194,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4691652640951073e-05,
+      "loss": 3.9017,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683266693440553e-05,
+      "loss": 3.9046,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4674897124733765e-05,
+      "loss": 3.9033,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666511177223245e-05,
+      "loss": 3.9113,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465812522971272e-05,
+      "loss": 3.9211,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46497392822022e-05,
+      "loss": 3.9107,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464135333469168e-05,
+      "loss": 3.9065,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4632967387181165e-05,
+      "loss": 3.9165,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4624581439670645e-05,
+      "loss": 3.9087,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4616195492160125e-05,
+      "loss": 3.9125,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46078095446496e-05,
+      "loss": 3.9048,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459942359713908e-05,
+      "loss": 3.8884,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4591054028432294e-05,
+      "loss": 3.9341,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4582668080921774e-05,
+      "loss": 3.919,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4574282133411254e-05,
+      "loss": 3.9158,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4565896185900734e-05,
+      "loss": 3.9124,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557510238390214e-05,
+      "loss": 3.9122,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549140669683426e-05,
+      "loss": 3.9051,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4540754722172903e-05,
+      "loss": 3.9143,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532368774662383e-05,
+      "loss": 3.9044,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4523982827151866e-05,
+      "loss": 3.9154,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4515596879641346e-05,
+      "loss": 3.927,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4507227310934555e-05,
+      "loss": 3.9064,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4498841363424035e-05,
+      "loss": 3.8985,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490455415913515e-05,
+      "loss": 3.925,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482069468402995e-05,
+      "loss": 3.8918,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4473683520892475e-05,
+      "loss": 3.9075,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4465297573381958e-05,
+      "loss": 3.9158,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4456928004675167e-05,
+      "loss": 3.9159,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4448542057164647e-05,
+      "loss": 3.9092,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440156109654127e-05,
+      "loss": 3.9036,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4431770162143607e-05,
+      "loss": 3.8913,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4423384214633084e-05,
+      "loss": 3.903,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4414998267122567e-05,
+      "loss": 3.9163,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.440662869841578e-05,
+      "loss": 3.9053,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4398242750905256e-05,
+      "loss": 3.9141,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4389856803394736e-05,
+      "loss": 3.9171,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381470855884216e-05,
+      "loss": 3.924,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373084908373696e-05,
+      "loss": 3.9058,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364698960863176e-05,
+      "loss": 3.9057,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.435632939215639e-05,
+      "loss": 3.9179,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.434794344464587e-05,
+      "loss": 3.9025,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433955749713535e-05,
+      "loss": 3.921,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433117154962483e-05,
+      "loss": 3.9086,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4322801980918037e-05,
+      "loss": 3.9075,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.431441603340752e-05,
+      "loss": 3.917,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4306030085897e-05,
+      "loss": 3.8957,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.429764413838648e-05,
+      "loss": 3.9055,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428927456967969e-05,
+      "loss": 3.9054,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428088862216917e-05,
+      "loss": 3.9175,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.427250267465865e-05,
+      "loss": 3.9042,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.426411672714813e-05,
+      "loss": 3.9107,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4255747158441342e-05,
+      "loss": 3.902,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247361210930822e-05,
+      "loss": 3.9036,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4238975263420302e-05,
+      "loss": 3.8983,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4230589315909782e-05,
+      "loss": 3.9108,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222203368399262e-05,
+      "loss": 3.921,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4213833799692474e-05,
+      "loss": 3.9143,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205447852181954e-05,
+      "loss": 3.915,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197061904671434e-05,
+      "loss": 3.8987,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418867595716091e-05,
+      "loss": 3.9062,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4180306388454123e-05,
+      "loss": 3.9078,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4171920440943603e-05,
+      "loss": 3.9153,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4163534493433083e-05,
+      "loss": 3.9148,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155148545922563e-05,
+      "loss": 3.9067,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4146762598412043e-05,
+      "loss": 3.9045,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138376650901523e-05,
+      "loss": 3.923,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4129990703391003e-05,
+      "loss": 3.8969,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4121604755880483e-05,
+      "loss": 3.9096,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113251565977428e-05,
+      "loss": 3.903,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4104865618466908e-05,
+      "loss": 3.9102,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4096479670956385e-05,
+      "loss": 3.9007,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4088093723445865e-05,
+      "loss": 3.9108,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079707775935345e-05,
+      "loss": 3.8982,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071321828424825e-05,
+      "loss": 3.9029,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4062935880914304e-05,
+      "loss": 3.907,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4054549933403784e-05,
+      "loss": 3.9039,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.404619674350073e-05,
+      "loss": 3.9187,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4037810795990206e-05,
+      "loss": 3.9232,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029424848479686e-05,
+      "loss": 3.9127,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021038900969166e-05,
+      "loss": 3.9102,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012652953458646e-05,
+      "loss": 3.8998,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.400426700594813e-05,
+      "loss": 3.9058,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.399588105843761e-05,
+      "loss": 3.9049,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987511489730818e-05,
+      "loss": 3.9213,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3979125542220298e-05,
+      "loss": 3.8902,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3970739594709778e-05,
+      "loss": 3.9091,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962353647199258e-05,
+      "loss": 3.9079,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3953967699688738e-05,
+      "loss": 3.9139,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3945581752178218e-05,
+      "loss": 3.9068,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3937195804667698e-05,
+      "loss": 3.9098,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392882623596091e-05,
+      "loss": 3.9006,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392045666725412e-05,
+      "loss": 3.9155,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.39120707197436e-05,
+      "loss": 3.9135,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3903684772233083e-05,
+      "loss": 3.9155,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895298824722563e-05,
+      "loss": 3.8994,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388691287721204e-05,
+      "loss": 3.9153,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387852692970152e-05,
+      "loss": 3.9165,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3870140982191e-05,
+      "loss": 3.907,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.386175503468048e-05,
+      "loss": 3.9078,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.385336908716996e-05,
+      "loss": 3.9191,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.384498313965944e-05,
+      "loss": 3.9069,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.383661357095265e-05,
+      "loss": 3.9299,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382822762344213e-05,
+      "loss": 3.9022,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381984167593161e-05,
+      "loss": 3.9139,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381145572842109e-05,
+      "loss": 3.904,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3803069780910568e-05,
+      "loss": 3.909,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794700212203784e-05,
+      "loss": 3.9103,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786314264693264e-05,
+      "loss": 3.9042,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3777928317182744e-05,
+      "loss": 3.9063,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3769558748475953e-05,
+      "loss": 3.8981,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3761172800965433e-05,
+      "loss": 3.9108,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.998835325241089,
+      "eval_runtime": 296.7069,
+      "eval_samples_per_second": 1286.088,
+      "eval_steps_per_second": 40.191,
+      "step": 1602720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3752786853454913e-05,
+      "loss": 3.9144,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3744400905944393e-05,
+      "loss": 3.9021,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3736031337237605e-05,
+      "loss": 3.9153,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3727645389727085e-05,
+      "loss": 3.9117,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3719259442216565e-05,
+      "loss": 3.9208,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3710873494706042e-05,
+      "loss": 3.9073,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3702487547195522e-05,
+      "loss": 3.9088,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3694101599685005e-05,
+      "loss": 3.8964,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3685715652174485e-05,
+      "loss": 3.9096,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3677329704663965e-05,
+      "loss": 3.9127,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3668943757153445e-05,
+      "loss": 3.9097,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3660557809642925e-05,
+      "loss": 3.9069,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3652188240936134e-05,
+      "loss": 3.9073,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3643802293425614e-05,
+      "loss": 3.9084,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3635416345915097e-05,
+      "loss": 3.9024,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3627030398404577e-05,
+      "loss": 3.8934,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3618644450894054e-05,
+      "loss": 3.9013,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3610258503383534e-05,
+      "loss": 3.91,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3601888934676746e-05,
+      "loss": 3.9045,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.359351936596996e-05,
+      "loss": 3.923,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.358513341845944e-05,
+      "loss": 3.9115,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.357674747094892e-05,
+      "loss": 3.9156,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.35683615234384e-05,
+      "loss": 3.9037,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3559975575927875e-05,
+      "loss": 3.9103,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3551589628417355e-05,
+      "loss": 3.9052,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3543203680906835e-05,
+      "loss": 3.9163,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3534817733396315e-05,
+      "loss": 3.9036,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.35264317858858e-05,
+      "loss": 3.9026,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.351804583837528e-05,
+      "loss": 3.8998,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3509659890864755e-05,
+      "loss": 3.9092,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3501273943354235e-05,
+      "loss": 3.9091,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3492920753451177e-05,
+      "loss": 3.9114,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.348453480594066e-05,
+      "loss": 3.9148,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.347614885843014e-05,
+      "loss": 3.91,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.346776291091962e-05,
+      "loss": 3.9067,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345939334221283e-05,
+      "loss": 3.9155,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345100739470231e-05,
+      "loss": 3.9105,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.344262144719179e-05,
+      "loss": 3.8968,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.343423549968127e-05,
+      "loss": 3.8976,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.342586593097448e-05,
+      "loss": 3.8946,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.341747998346396e-05,
+      "loss": 3.8997,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.340909403595344e-05,
+      "loss": 3.913,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.340070808844292e-05,
+      "loss": 3.9054,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.33923221409324e-05,
+      "loss": 3.8992,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3383952572225614e-05,
+      "loss": 3.9106,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3375566624715094e-05,
+      "loss": 3.9012,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3367180677204573e-05,
+      "loss": 3.9069,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3358794729694053e-05,
+      "loss": 3.8964,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.335040878218353e-05,
+      "loss": 3.8796,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.334202283467301e-05,
+      "loss": 3.9282,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.333363688716249e-05,
+      "loss": 3.9091,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.332525093965197e-05,
+      "loss": 3.9122,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3316864992141453e-05,
+      "loss": 3.9013,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3308495423434662e-05,
+      "loss": 3.9103,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3300109475924142e-05,
+      "loss": 3.897,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3291723528413622e-05,
+      "loss": 3.9031,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3283337580903102e-05,
+      "loss": 3.9026,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3274951633392582e-05,
+      "loss": 3.9043,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3266565685882062e-05,
+      "loss": 3.9229,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3258196117175275e-05,
+      "loss": 3.8975,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3249810169664755e-05,
+      "loss": 3.8902,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3241424222154234e-05,
+      "loss": 3.9172,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.323303827464371e-05,
+      "loss": 3.8854,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3224668705936924e-05,
+      "loss": 3.8939,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3216282758426407e-05,
+      "loss": 3.9116,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3207896810915883e-05,
+      "loss": 3.9099,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3199510863405363e-05,
+      "loss": 3.8981,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3191141294698576e-05,
+      "loss": 3.9026,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3182755347188056e-05,
+      "loss": 3.8811,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3174369399677532e-05,
+      "loss": 3.8993,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3165983452167016e-05,
+      "loss": 3.9076,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3157613883460228e-05,
+      "loss": 3.8939,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3149227935949708e-05,
+      "loss": 3.9081,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3140841988439185e-05,
+      "loss": 3.9084,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3132456040928665e-05,
+      "loss": 3.9196,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3124070093418145e-05,
+      "loss": 3.8947,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3115684145907625e-05,
+      "loss": 3.9025,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3107298198397108e-05,
+      "loss": 3.9123,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3098912250886588e-05,
+      "loss": 3.895,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3090542682179797e-05,
+      "loss": 3.9094,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3082156734669277e-05,
+      "loss": 3.9028,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3073770787158757e-05,
+      "loss": 3.899,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3065384839648237e-05,
+      "loss": 3.9118,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3056998892137717e-05,
+      "loss": 3.8912,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304862932343093e-05,
+      "loss": 3.8944,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304024337592041e-05,
+      "loss": 3.9009,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.303185742840989e-05,
+      "loss": 3.9099,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3023471480899366e-05,
+      "loss": 3.9001,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3015101912192578e-05,
+      "loss": 3.9027,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.300671596468206e-05,
+      "loss": 3.8901,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2998330017171538e-05,
+      "loss": 3.9002,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2989944069661018e-05,
+      "loss": 3.8925,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2981558122150498e-05,
+      "loss": 3.9007,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2973172174639978e-05,
+      "loss": 3.9177,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.296480260593319e-05,
+      "loss": 3.9046,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.295641665842267e-05,
+      "loss": 3.9064,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.294803071091215e-05,
+      "loss": 3.893,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.293964476340163e-05,
+      "loss": 3.8988,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.293125881589111e-05,
+      "loss": 3.8994,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.292287286838059e-05,
+      "loss": 3.91,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2914486920870067e-05,
+      "loss": 3.9022,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2906100973359547e-05,
+      "loss": 3.9073,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2897731404652763e-05,
+      "loss": 3.8967,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2889345457142243e-05,
+      "loss": 3.9132,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2880975888435452e-05,
+      "loss": 3.8933,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2872589940924932e-05,
+      "loss": 3.9038,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.286420399341441e-05,
+      "loss": 3.8917,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.285581804590389e-05,
+      "loss": 3.9072,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.284743209839337e-05,
+      "loss": 3.8949,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.283904615088285e-05,
+      "loss": 3.8993,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2830676582176064e-05,
+      "loss": 3.8931,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.282229063466554e-05,
+      "loss": 3.8992,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.281390468715502e-05,
+      "loss": 3.8974,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.28055187396445e-05,
+      "loss": 3.8988,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2797132792133984e-05,
+      "loss": 3.9095,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2788746844623464e-05,
+      "loss": 3.915,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2780360897112944e-05,
+      "loss": 3.9054,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2771974949602424e-05,
+      "loss": 3.9084,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.27635890020919e-05,
+      "loss": 3.8907,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.275520305458138e-05,
+      "loss": 3.8986,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.274681710707086e-05,
+      "loss": 3.9029,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.273843115956034e-05,
+      "loss": 3.9125,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2730061590853553e-05,
+      "loss": 3.884,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2721675643343033e-05,
+      "loss": 3.9041,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2713289695832513e-05,
+      "loss": 3.9029,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.270492012712572e-05,
+      "loss": 3.9031,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.26965341796152e-05,
+      "loss": 3.8988,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2688164610908417e-05,
+      "loss": 3.9036,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2679778663397897e-05,
+      "loss": 3.8957,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2671392715887374e-05,
+      "loss": 3.9083,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2663006768376854e-05,
+      "loss": 3.9075,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2654620820866334e-05,
+      "loss": 3.909,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2646234873355814e-05,
+      "loss": 3.8925,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2637848925845294e-05,
+      "loss": 3.9074,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2629462978334777e-05,
+      "loss": 3.911,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2621093409627986e-05,
+      "loss": 3.8995,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2612707462117466e-05,
+      "loss": 3.9035,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2604321514606946e-05,
+      "loss": 3.9092,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2595935567096426e-05,
+      "loss": 3.9065,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2587549619585903e-05,
+      "loss": 3.9158,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.257918005087912e-05,
+      "loss": 3.8994,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.25707941033686e-05,
+      "loss": 3.9032,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.256240815585808e-05,
+      "loss": 3.8997,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2554022208347555e-05,
+      "loss": 3.9029,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.25456690184445e-05,
+      "loss": 3.8981,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.253728307093398e-05,
+      "loss": 3.8947,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252889712342346e-05,
+      "loss": 3.9034,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252052755471667e-05,
+      "loss": 3.8889,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.251214160720615e-05,
+      "loss": 3.9075,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.995767116546631,
+      "eval_runtime": 295.7244,
+      "eval_samples_per_second": 1290.36,
+      "eval_steps_per_second": 40.325,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.250375565969563e-05,
+      "loss": 3.9105,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.249536971218511e-05,
+      "loss": 3.8964,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2486983764674592e-05,
+      "loss": 3.9074,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2478597817164072e-05,
+      "loss": 3.9046,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2470211869653552e-05,
+      "loss": 3.9145,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246182592214303e-05,
+      "loss": 3.9018,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.245343997463251e-05,
+      "loss": 3.903,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.244505402712199e-05,
+      "loss": 3.89,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243666807961147e-05,
+      "loss": 3.9036,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242828213210095e-05,
+      "loss": 3.9047,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2419896184590432e-05,
+      "loss": 3.9013,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2411510237079912e-05,
+      "loss": 3.8997,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.240314066837312e-05,
+      "loss": 3.8982,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23947547208626e-05,
+      "loss": 3.9027,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238636877335208e-05,
+      "loss": 3.8977,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2377982825841557e-05,
+      "loss": 3.8869,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.236959687833104e-05,
+      "loss": 3.8968,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2361227309624253e-05,
+      "loss": 3.9003,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2352841362113733e-05,
+      "loss": 3.9005,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234445541460321e-05,
+      "loss": 3.9114,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233606946709269e-05,
+      "loss": 3.9055,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232768351958217e-05,
+      "loss": 3.9115,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231929757207165e-05,
+      "loss": 3.8941,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2310911624561133e-05,
+      "loss": 3.9042,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2302525677050613e-05,
+      "loss": 3.8994,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229413972954009e-05,
+      "loss": 3.9086,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.228575378202957e-05,
+      "loss": 3.8965,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227736783451905e-05,
+      "loss": 3.899,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2268998265812262e-05,
+      "loss": 3.8935,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2260612318301742e-05,
+      "loss": 3.8984,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252226370791222e-05,
+      "loss": 3.8994,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2243840423280702e-05,
+      "loss": 3.9081,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235454475770182e-05,
+      "loss": 3.9068,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2227068528259662e-05,
+      "loss": 3.9042,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.221869895955287e-05,
+      "loss": 3.8988,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210329390846087e-05,
+      "loss": 3.904,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2201943443335563e-05,
+      "loss": 3.9052,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193557495825043e-05,
+      "loss": 3.8929,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185171548314523e-05,
+      "loss": 3.8888,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2176801979607736e-05,
+      "loss": 3.8863,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168416032097216e-05,
+      "loss": 3.896,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2160030084586696e-05,
+      "loss": 3.9011,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151644137076176e-05,
+      "loss": 3.8991,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143258189565655e-05,
+      "loss": 3.8951,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2134872242055135e-05,
+      "loss": 3.901,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126486294544615e-05,
+      "loss": 3.8969,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118100347034095e-05,
+      "loss": 3.9011,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109730778327304e-05,
+      "loss": 3.8896,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101344830816788e-05,
+      "loss": 3.8745,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2092958883306268e-05,
+      "loss": 3.9175,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084572935795744e-05,
+      "loss": 3.9007,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076186988285224e-05,
+      "loss": 3.9068,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2067817419578437e-05,
+      "loss": 3.8957,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2059431472067917e-05,
+      "loss": 3.9007,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2051045524557397e-05,
+      "loss": 3.8911,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2042659577046877e-05,
+      "loss": 3.8944,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2034273629536357e-05,
+      "loss": 3.8996,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2025887682025837e-05,
+      "loss": 3.8938,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017501734515316e-05,
+      "loss": 3.9165,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2009132165808526e-05,
+      "loss": 3.8956,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200074621829801e-05,
+      "loss": 3.8858,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199236027078749e-05,
+      "loss": 3.9103,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.198397432327697e-05,
+      "loss": 3.8779,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197558837576645e-05,
+      "loss": 3.8832,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1967218807059658e-05,
+      "loss": 3.9091,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1958832859549138e-05,
+      "loss": 3.904,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1950446912038618e-05,
+      "loss": 3.8896,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19420609645281e-05,
+      "loss": 3.8965,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1933675017017578e-05,
+      "loss": 3.877,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1925289069507058e-05,
+      "loss": 3.8939,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1916903121996538e-05,
+      "loss": 3.8967,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1908517174486018e-05,
+      "loss": 3.8924,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1900131226975498e-05,
+      "loss": 3.8986,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1891745279464977e-05,
+      "loss": 3.902,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188337571075819e-05,
+      "loss": 3.9112,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187498976324767e-05,
+      "loss": 3.89,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186662019454088e-05,
+      "loss": 3.8962,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.185823424703036e-05,
+      "loss": 3.9023,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184984829951984e-05,
+      "loss": 3.8916,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184146235200932e-05,
+      "loss": 3.9019,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.183309278330253e-05,
+      "loss": 3.8976,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.182470683579201e-05,
+      "loss": 3.8912,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1816337267085224e-05,
+      "loss": 3.9114,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18079513195747e-05,
+      "loss": 3.8824,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.179956537206418e-05,
+      "loss": 3.8868,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1791179424553664e-05,
+      "loss": 3.8976,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1782809855846873e-05,
+      "loss": 3.9009,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774423908336353e-05,
+      "loss": 3.8953,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766037960825833e-05,
+      "loss": 3.8962,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757652013315313e-05,
+      "loss": 3.8819,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1749266065804793e-05,
+      "loss": 3.8931,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1740880118294273e-05,
+      "loss": 3.8872,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732494170783756e-05,
+      "loss": 3.8956,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724108223273232e-05,
+      "loss": 3.91,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715722275762712e-05,
+      "loss": 3.8985,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707336328252192e-05,
+      "loss": 3.8998,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.16989667595454e-05,
+      "loss": 3.8882,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169058081203488e-05,
+      "loss": 3.8896,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682194864524365e-05,
+      "loss": 3.8895,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1673808917013845e-05,
+      "loss": 3.9042,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665422969503325e-05,
+      "loss": 3.8995,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657053400796534e-05,
+      "loss": 3.9008,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648683832089746e-05,
+      "loss": 3.8893,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1640297884579226e-05,
+      "loss": 3.9028,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1631911937068706e-05,
+      "loss": 3.8904,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1623525989558186e-05,
+      "loss": 3.8968,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615140042047666e-05,
+      "loss": 3.8815,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606754094537146e-05,
+      "loss": 3.8999,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598368147026626e-05,
+      "loss": 3.8885,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1589982199516106e-05,
+      "loss": 3.8943,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581596252005586e-05,
+      "loss": 3.8832,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.15732266832988e-05,
+      "loss": 3.8976,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.156484073578828e-05,
+      "loss": 3.8881,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1556471167081487e-05,
+      "loss": 3.8906,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1548085219570967e-05,
+      "loss": 3.9041,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1539699272060447e-05,
+      "loss": 3.9088,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1531313324549927e-05,
+      "loss": 3.9031,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.152292737703941e-05,
+      "loss": 3.8953,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1514541429528887e-05,
+      "loss": 3.8882,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1506155482018367e-05,
+      "loss": 3.8907,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.149778591331158e-05,
+      "loss": 3.8975,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489399965801056e-05,
+      "loss": 3.9013,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1481014018290536e-05,
+      "loss": 3.8828,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.147262807078002e-05,
+      "loss": 3.8968,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.14642421232695e-05,
+      "loss": 3.8969,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.145585617575898e-05,
+      "loss": 3.8962,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.144748660705219e-05,
+      "loss": 3.8946,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.143910065954167e-05,
+      "loss": 3.8955,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.143071471203115e-05,
+      "loss": 3.8872,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.142232876452063e-05,
+      "loss": 3.9031,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1413942817010112e-05,
+      "loss": 3.9029,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.140555686949959e-05,
+      "loss": 3.8998,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1397170921989068e-05,
+      "loss": 3.8866,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1388784974478548e-05,
+      "loss": 3.9043,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138041540577176e-05,
+      "loss": 3.9005,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.137202945826124e-05,
+      "loss": 3.894,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.136364351075072e-05,
+      "loss": 3.898,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.13552575632402e-05,
+      "loss": 3.8999,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.134687161572968e-05,
+      "loss": 3.8983,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133850204702289e-05,
+      "loss": 3.9103,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133011609951237e-05,
+      "loss": 3.8906,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.132173015200185e-05,
+      "loss": 3.8999,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.131334420449133e-05,
+      "loss": 3.8973,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1304958256980813e-05,
+      "loss": 3.8956,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1296572309470293e-05,
+      "loss": 3.8889,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288202740763502e-05,
+      "loss": 3.8885,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1279816793252982e-05,
+      "loss": 3.8939,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1271447224546194e-05,
+      "loss": 3.8868,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1263061277035674e-05,
+      "loss": 3.9004,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.993607997894287,
+      "eval_runtime": 294.1705,
+      "eval_samples_per_second": 1297.176,
+      "eval_steps_per_second": 40.538,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1254675329525154e-05,
+      "loss": 3.8988,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1246289382014634e-05,
+      "loss": 3.8905,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1237903434504114e-05,
+      "loss": 3.897,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1229517486993594e-05,
+      "loss": 3.898,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.122113153948307e-05,
+      "loss": 3.9061,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.121274559197255e-05,
+      "loss": 3.8952,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1204376023265766e-05,
+      "loss": 3.9004,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1195990075755243e-05,
+      "loss": 3.8875,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1187604128244723e-05,
+      "loss": 3.8897,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1179218180734203e-05,
+      "loss": 3.9036,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1170832233223683e-05,
+      "loss": 3.8952,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1162446285713163e-05,
+      "loss": 3.8911,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1154076717006375e-05,
+      "loss": 3.9012,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1145690769495855e-05,
+      "loss": 3.8893,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1137304821985335e-05,
+      "loss": 3.8874,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1128918874474815e-05,
+      "loss": 3.8828,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1120532926964295e-05,
+      "loss": 3.8883,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1112146979453775e-05,
+      "loss": 3.8941,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1103761031943252e-05,
+      "loss": 3.8875,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1095375084432735e-05,
+      "loss": 3.9096,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1086989136922215e-05,
+      "loss": 3.9008,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1078603189411695e-05,
+      "loss": 3.9038,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1070217241901175e-05,
+      "loss": 3.891,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1061831294390655e-05,
+      "loss": 3.8941,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1053461725683864e-05,
+      "loss": 3.9008,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1045075778173344e-05,
+      "loss": 3.8994,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1036689830662827e-05,
+      "loss": 3.8909,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1028303883152307e-05,
+      "loss": 3.895,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1019934314445516e-05,
+      "loss": 3.8858,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1011548366934996e-05,
+      "loss": 3.8932,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1003178798228205e-05,
+      "loss": 3.8939,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.099479285071769e-05,
+      "loss": 3.9008,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0986423282010898e-05,
+      "loss": 3.8994,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.097805371330411e-05,
+      "loss": 3.8994,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.096966776579359e-05,
+      "loss": 3.8938,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.096128181828307e-05,
+      "loss": 3.8954,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.095289587077255e-05,
+      "loss": 3.9,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.094450992326203e-05,
+      "loss": 3.8859,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.093612397575151e-05,
+      "loss": 3.8822,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.092773802824099e-05,
+      "loss": 3.8815,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.091935208073047e-05,
+      "loss": 3.8891,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.091096613321995e-05,
+      "loss": 3.8908,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.090258018570943e-05,
+      "loss": 3.8947,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894210617002642e-05,
+      "loss": 3.8907,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0885824669492122e-05,
+      "loss": 3.8912,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877438721981602e-05,
+      "loss": 3.8922,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.086905277447108e-05,
+      "loss": 3.8975,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.086066682696056e-05,
+      "loss": 3.8824,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.085229725825377e-05,
+      "loss": 3.867,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.084391131074325e-05,
+      "loss": 3.9058,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.083552536323273e-05,
+      "loss": 3.8975,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.082713941572221e-05,
+      "loss": 3.9037,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.081875346821169e-05,
+      "loss": 3.8885,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.081036752070117e-05,
+      "loss": 3.8919,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.080198157319065e-05,
+      "loss": 3.886,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.079359562568013e-05,
+      "loss": 3.8881,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.078520967816961e-05,
+      "loss": 3.8968,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.077682373065909e-05,
+      "loss": 3.8818,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076843778314857e-05,
+      "loss": 3.912,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076005183563805e-05,
+      "loss": 3.8916,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0751698645734992e-05,
+      "loss": 3.8739,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0743312698224472e-05,
+      "loss": 3.9044,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0734926750713952e-05,
+      "loss": 3.8784,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0726540803203436e-05,
+      "loss": 3.8752,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0718171234496645e-05,
+      "loss": 3.9029,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0709785286986125e-05,
+      "loss": 3.8956,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0701399339475605e-05,
+      "loss": 3.8845,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0693013391965085e-05,
+      "loss": 3.8889,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.068462744445456e-05,
+      "loss": 3.8728,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0676257875747777e-05,
+      "loss": 3.8863,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0667871928237257e-05,
+      "loss": 3.8918,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659485980726734e-05,
+      "loss": 3.8854,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651100033216214e-05,
+      "loss": 3.8925,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0642730464509426e-05,
+      "loss": 3.8927,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0634344516998906e-05,
+      "loss": 3.9043,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.062597494829212e-05,
+      "loss": 3.8836,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.06175890007816e-05,
+      "loss": 3.8918,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060920305327108e-05,
+      "loss": 3.8946,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060081710576056e-05,
+      "loss": 3.8844,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0592431158250035e-05,
+      "loss": 3.8971,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.058406158954325e-05,
+      "loss": 3.8927,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.057567564203273e-05,
+      "loss": 3.8826,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0567289694522207e-05,
+      "loss": 3.904,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0558903747011687e-05,
+      "loss": 3.8739,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0550517799501167e-05,
+      "loss": 3.8805,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0542131851990647e-05,
+      "loss": 3.8934,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0533745904480127e-05,
+      "loss": 3.8931,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0525359956969607e-05,
+      "loss": 3.8861,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.051697400945909e-05,
+      "loss": 3.8939,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0508588061948567e-05,
+      "loss": 3.8755,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0500202114438047e-05,
+      "loss": 3.8829,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.049183254573126e-05,
+      "loss": 3.8847,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0483446598220736e-05,
+      "loss": 3.8853,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.047506065071022e-05,
+      "loss": 3.9051,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.04666747031997e-05,
+      "loss": 3.8914,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.045828875568918e-05,
+      "loss": 3.8967,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.044990280817866e-05,
+      "loss": 3.8831,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.044151686066814e-05,
+      "loss": 3.8846,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.043313091315762e-05,
+      "loss": 3.8864,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0424761344450828e-05,
+      "loss": 3.8949,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416375396940308e-05,
+      "loss": 3.8918,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.040800582823352e-05,
+      "loss": 3.8945,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0399619880723e-05,
+      "loss": 3.8861,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.039123393321248e-05,
+      "loss": 3.8921,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.038284798570196e-05,
+      "loss": 3.8888,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.037446203819144e-05,
+      "loss": 3.8889,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.036607609068092e-05,
+      "loss": 3.8768,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.03576901431704e-05,
+      "loss": 3.8976,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.034930419565988e-05,
+      "loss": 3.8807,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.034091824814936e-05,
+      "loss": 3.8869,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.033253230063884e-05,
+      "loss": 3.8777,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.032414635312832e-05,
+      "loss": 3.8881,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.03157604056178e-05,
+      "loss": 3.8845,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0307374458107277e-05,
+      "loss": 3.8851,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0299004889400493e-05,
+      "loss": 3.8965,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0290618941889973e-05,
+      "loss": 3.9026,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0282232994379453e-05,
+      "loss": 3.9001,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.027386342567266e-05,
+      "loss": 3.8876,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.026547747816214e-05,
+      "loss": 3.8814,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025709153065162e-05,
+      "loss": 3.8843,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.02487055831411e-05,
+      "loss": 3.8913,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.024031963563058e-05,
+      "loss": 3.8931,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.023193368812006e-05,
+      "loss": 3.875,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0223564119413274e-05,
+      "loss": 3.8911,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021517817190275e-05,
+      "loss": 3.8923,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.020679222439223e-05,
+      "loss": 3.8868,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0198406276881714e-05,
+      "loss": 3.8878,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0190036708174926e-05,
+      "loss": 3.892,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0181650760664403e-05,
+      "loss": 3.8836,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0173264813153883e-05,
+      "loss": 3.8953,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0164878865643363e-05,
+      "loss": 3.8928,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0156492918132843e-05,
+      "loss": 3.8983,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0148106970622323e-05,
+      "loss": 3.8789,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0139737401915535e-05,
+      "loss": 3.8956,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0131351454405015e-05,
+      "loss": 3.8956,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0122965506894495e-05,
+      "loss": 3.8891,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0114579559383975e-05,
+      "loss": 3.8933,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0106193611873455e-05,
+      "loss": 3.8914,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0097824043166667e-05,
+      "loss": 3.892,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0089438095656147e-05,
+      "loss": 3.9076,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0081068526949357e-05,
+      "loss": 3.8819,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0072682579438836e-05,
+      "loss": 3.8951,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0064296631928316e-05,
+      "loss": 3.8924,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0055910684417796e-05,
+      "loss": 3.8911,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0047524736907276e-05,
+      "loss": 3.8843,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0039138789396756e-05,
+      "loss": 3.8811,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0030752841886236e-05,
+      "loss": 3.891,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.002238327317945e-05,
+      "loss": 3.8782,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.001399732566893e-05,
+      "loss": 3.8954,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.991264581680298,
+      "eval_runtime": 294.8059,
+      "eval_samples_per_second": 1294.38,
+      "eval_steps_per_second": 40.45,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0005611378158405e-05,
+      "loss": 3.8897,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9997225430647885e-05,
+      "loss": 3.8872,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998883948313737e-05,
+      "loss": 3.8916,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998045353562685e-05,
+      "loss": 3.8938,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997206758811633e-05,
+      "loss": 3.8994,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996368164060581e-05,
+      "loss": 3.8907,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9955295693095285e-05,
+      "loss": 3.8941,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9946909745584765e-05,
+      "loss": 3.8787,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9938523798074245e-05,
+      "loss": 3.8857,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9930137850563725e-05,
+      "loss": 3.897,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9921751903053208e-05,
+      "loss": 3.8906,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9913365955542688e-05,
+      "loss": 3.8856,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9904996386835897e-05,
+      "loss": 3.8977,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896610439325377e-05,
+      "loss": 3.883,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888224491814857e-05,
+      "loss": 3.8826,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9879838544304337e-05,
+      "loss": 3.8738,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9871452596793817e-05,
+      "loss": 3.8857,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9863066649283297e-05,
+      "loss": 3.894,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9854680701772777e-05,
+      "loss": 3.8821,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9846294754262257e-05,
+      "loss": 3.9005,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9837925185555466e-05,
+      "loss": 3.8953,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829539238044946e-05,
+      "loss": 3.8965,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.982115329053443e-05,
+      "loss": 3.8893,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.981276734302391e-05,
+      "loss": 3.8906,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.980438139551339e-05,
+      "loss": 3.8914,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.979599544800287e-05,
+      "loss": 3.8909,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.978760950049235e-05,
+      "loss": 3.882,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779223552981826e-05,
+      "loss": 3.8907,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9770853984275038e-05,
+      "loss": 3.8805,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.976248441556825e-05,
+      "loss": 3.8864,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.975409846805773e-05,
+      "loss": 3.8888,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.974571252054721e-05,
+      "loss": 3.8904,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.973732657303669e-05,
+      "loss": 3.897,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.972894062552617e-05,
+      "loss": 3.8978,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720554678015647e-05,
+      "loss": 3.8837,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9712185109308863e-05,
+      "loss": 3.8934,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9703815540602072e-05,
+      "loss": 3.8957,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9695429593091552e-05,
+      "loss": 3.8772,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9687043645581032e-05,
+      "loss": 3.8801,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9678657698070512e-05,
+      "loss": 3.877,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9670271750559992e-05,
+      "loss": 3.8794,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9661885803049472e-05,
+      "loss": 3.8872,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653499855538952e-05,
+      "loss": 3.8906,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9645113908028432e-05,
+      "loss": 3.8855,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.963672796051791e-05,
+      "loss": 3.8842,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.962834201300739e-05,
+      "loss": 3.8855,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.96199724443006e-05,
+      "loss": 3.8953,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9611586496790084e-05,
+      "loss": 3.8748,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9603200549279564e-05,
+      "loss": 3.8651,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9594830980572773e-05,
+      "loss": 3.8972,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586445033062253e-05,
+      "loss": 3.8905,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9578059085551733e-05,
+      "loss": 3.8972,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9569673138041213e-05,
+      "loss": 3.8827,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9561287190530693e-05,
+      "loss": 3.8853,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9552901243020176e-05,
+      "loss": 3.8855,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9544515295509656e-05,
+      "loss": 3.8823,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9536129347999133e-05,
+      "loss": 3.8935,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9527759779292345e-05,
+      "loss": 3.8757,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9519373831781825e-05,
+      "loss": 3.909,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9510987884271302e-05,
+      "loss": 3.8835,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9502601936760785e-05,
+      "loss": 3.8675,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9494232368053998e-05,
+      "loss": 3.8992,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9485846420543478e-05,
+      "loss": 3.8769,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9477460473032954e-05,
+      "loss": 3.868,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9469074525522434e-05,
+      "loss": 3.8945,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9460704956815647e-05,
+      "loss": 3.8912,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945231900930513e-05,
+      "loss": 3.8822,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9443933061794607e-05,
+      "loss": 3.8822,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9435547114284086e-05,
+      "loss": 3.866,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9427161166773566e-05,
+      "loss": 3.8773,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9418775219263046e-05,
+      "loss": 3.8892,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9410389271752526e-05,
+      "loss": 3.8803,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9402003324242006e-05,
+      "loss": 3.8844,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939363375553522e-05,
+      "loss": 3.8842,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.93852478080247e-05,
+      "loss": 3.8993,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9376878239317908e-05,
+      "loss": 3.8818,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9368492291807388e-05,
+      "loss": 3.8851,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9360106344296868e-05,
+      "loss": 3.8864,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351720396786348e-05,
+      "loss": 3.8815,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.934333444927583e-05,
+      "loss": 3.8914,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9334948501765308e-05,
+      "loss": 3.8844,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.932657893305852e-05,
+      "loss": 3.8781,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9318192985548e-05,
+      "loss": 3.8946,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.930980703803748e-05,
+      "loss": 3.8711,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301421090526957e-05,
+      "loss": 3.8753,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.929303514301644e-05,
+      "loss": 3.8849,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.928468195311338e-05,
+      "loss": 3.8903,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.927629600560286e-05,
+      "loss": 3.8775,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.926791005809234e-05,
+      "loss": 3.8903,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.925952411058182e-05,
+      "loss": 3.8734,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.92511381630713e-05,
+      "loss": 3.878,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242768594364514e-05,
+      "loss": 3.8742,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9234382646853994e-05,
+      "loss": 3.8807,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9225996699343474e-05,
+      "loss": 3.899,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9217610751832954e-05,
+      "loss": 3.8865,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.920922480432243e-05,
+      "loss": 3.8895,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.920083885681191e-05,
+      "loss": 3.8793,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9192452909301394e-05,
+      "loss": 3.8777,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9184066961790874e-05,
+      "loss": 3.8807,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9175681014280354e-05,
+      "loss": 3.8915,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167311445573563e-05,
+      "loss": 3.8863,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9158925498063043e-05,
+      "loss": 3.8861,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9150539550552523e-05,
+      "loss": 3.8864,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9142153603042002e-05,
+      "loss": 3.8828,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9133767655531486e-05,
+      "loss": 3.8814,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9125398086824695e-05,
+      "loss": 3.8814,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9117012139314175e-05,
+      "loss": 3.873,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9108626191803655e-05,
+      "loss": 3.8889,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9100240244293135e-05,
+      "loss": 3.8772,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9091854296782615e-05,
+      "loss": 3.8809,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9083468349272095e-05,
+      "loss": 3.878,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9075098780565307e-05,
+      "loss": 3.8801,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066712833054787e-05,
+      "loss": 3.8803,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058326885544264e-05,
+      "loss": 3.8789,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9049940938033744e-05,
+      "loss": 3.8955,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9041554990523224e-05,
+      "loss": 3.8938,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9033185421816436e-05,
+      "loss": 3.895,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9024799474305916e-05,
+      "loss": 3.8825,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9016413526795396e-05,
+      "loss": 3.88,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9008027579284876e-05,
+      "loss": 3.8772,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8999641631774356e-05,
+      "loss": 3.8857,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8991255684263836e-05,
+      "loss": 3.8905,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.898288611555705e-05,
+      "loss": 3.8728,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897450016804653e-05,
+      "loss": 3.8834,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8966114220536008e-05,
+      "loss": 3.8856,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8957728273025488e-05,
+      "loss": 3.8852,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8949342325514968e-05,
+      "loss": 3.8815,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894098913561191e-05,
+      "loss": 3.8846,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.893260318810139e-05,
+      "loss": 3.8764,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.892421724059087e-05,
+      "loss": 3.8916,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.891583129308035e-05,
+      "loss": 3.8851,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890744534556983e-05,
+      "loss": 3.8964,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889905939805931e-05,
+      "loss": 3.8749,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889067345054879e-05,
+      "loss": 3.8874,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.888228750303827e-05,
+      "loss": 3.893,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.887390155552775e-05,
+      "loss": 3.8829,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8865531986820962e-05,
+      "loss": 3.885,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8857146039310442e-05,
+      "loss": 3.8876,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.884876009179992e-05,
+      "loss": 3.8871,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.88403741442894e-05,
+      "loss": 3.9004,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.883198819677888e-05,
+      "loss": 3.8789,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.882360224926836e-05,
+      "loss": 3.8884,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.881523268056157e-05,
+      "loss": 3.887,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.880684673305105e-05,
+      "loss": 3.8822,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798477164344263e-05,
+      "loss": 3.8785,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.879009121683374e-05,
+      "loss": 3.8747,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8781705269323223e-05,
+      "loss": 3.8814,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773319321812703e-05,
+      "loss": 3.8807,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8764933374302183e-05,
+      "loss": 3.8852,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.989142894744873,
+      "eval_runtime": 298.6552,
+      "eval_samples_per_second": 1277.698,
+      "eval_steps_per_second": 39.929,
+      "step": 1908000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8756547426791663e-05,
+      "loss": 3.8841,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8748161479281143e-05,
+      "loss": 3.8815,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.873977553177062e-05,
+      "loss": 3.8829,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.87313895842601e-05,
+      "loss": 3.8901,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.872300363674958e-05,
+      "loss": 3.8964,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8714617689239063e-05,
+      "loss": 3.8867,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8706231741728543e-05,
+      "loss": 3.8864,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8697862173021752e-05,
+      "loss": 3.8769,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8689476225511232e-05,
+      "loss": 3.8806,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8681090278000712e-05,
+      "loss": 3.888,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8672704330490192e-05,
+      "loss": 3.8864,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8664334761783404e-05,
+      "loss": 3.8803,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8655965193076617e-05,
+      "loss": 3.892,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8647579245566093e-05,
+      "loss": 3.8775,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8639193298055573e-05,
+      "loss": 3.8766,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8630807350545053e-05,
+      "loss": 3.8661,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8622437781838266e-05,
+      "loss": 3.8846,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8614068213131478e-05,
+      "loss": 3.8887,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8605682265620958e-05,
+      "loss": 3.8746,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8597296318110438e-05,
+      "loss": 3.8936,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8588910370599918e-05,
+      "loss": 3.8891,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8580524423089395e-05,
+      "loss": 3.8904,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8572138475578878e-05,
+      "loss": 3.8846,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8563752528068358e-05,
+      "loss": 3.8856,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8555366580557838e-05,
+      "loss": 3.885,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8546980633047318e-05,
+      "loss": 3.8914,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8538611064340527e-05,
+      "loss": 3.8733,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8530225116830007e-05,
+      "loss": 3.887,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8521839169319487e-05,
+      "loss": 3.8732,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.851345322180897e-05,
+      "loss": 3.8822,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.850506727429845e-05,
+      "loss": 3.8862,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.849669770559166e-05,
+      "loss": 3.8869,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.848831175808114e-05,
+      "loss": 3.8839,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.847992581057062e-05,
+      "loss": 3.8956,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.84715398630601e-05,
+      "loss": 3.8804,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.846317029435331e-05,
+      "loss": 3.886,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.845478434684279e-05,
+      "loss": 3.8889,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.844639839933227e-05,
+      "loss": 3.8698,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8438012451821748e-05,
+      "loss": 3.8761,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8429626504311228e-05,
+      "loss": 3.876,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8421240556800708e-05,
+      "loss": 3.8708,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8412854609290188e-05,
+      "loss": 3.8794,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.840446866177967e-05,
+      "loss": 3.8894,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.839608271426915e-05,
+      "loss": 3.8758,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.838771314556236e-05,
+      "loss": 3.8809,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8379343576855573e-05,
+      "loss": 3.878,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.837095762934505e-05,
+      "loss": 3.8866,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8362571681834533e-05,
+      "loss": 3.8688,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8354185734324013e-05,
+      "loss": 3.8626,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8345799786813493e-05,
+      "loss": 3.8926,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8337413839302973e-05,
+      "loss": 3.8878,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8329027891792453e-05,
+      "loss": 3.8958,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.832064194428193e-05,
+      "loss": 3.8757,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.831227237557514e-05,
+      "loss": 3.8794,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8303886428064625e-05,
+      "loss": 3.8807,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8295500480554105e-05,
+      "loss": 3.8734,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.828711453304358e-05,
+      "loss": 3.8918,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.827872858553306e-05,
+      "loss": 3.863,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8270359016826274e-05,
+      "loss": 3.9047,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8261973069315754e-05,
+      "loss": 3.8823,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8253587121805234e-05,
+      "loss": 3.8601,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8245201174294714e-05,
+      "loss": 3.8889,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8236831605587926e-05,
+      "loss": 3.8752,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8228445658077403e-05,
+      "loss": 3.8626,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8220076089370615e-05,
+      "loss": 3.8873,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8211690141860095e-05,
+      "loss": 3.8895,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.820330419434958e-05,
+      "loss": 3.8738,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8194918246839055e-05,
+      "loss": 3.8794,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8186532299328535e-05,
+      "loss": 3.8592,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8178146351818015e-05,
+      "loss": 3.8739,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8169760404307495e-05,
+      "loss": 3.8806,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8161374456796975e-05,
+      "loss": 3.8727,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8152988509286455e-05,
+      "loss": 3.8842,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8144618940579667e-05,
+      "loss": 3.8751,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8136232993069147e-05,
+      "loss": 3.8958,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8127847045558627e-05,
+      "loss": 3.8767,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8119461098048107e-05,
+      "loss": 3.8816,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8111075150537584e-05,
+      "loss": 3.8805,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8102689203027064e-05,
+      "loss": 3.8741,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8094303255516544e-05,
+      "loss": 3.8843,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.808593368680976e-05,
+      "loss": 3.8776,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8077547739299236e-05,
+      "loss": 3.8801,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8069161791788716e-05,
+      "loss": 3.8852,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8060775844278196e-05,
+      "loss": 3.8727,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8052406275571405e-05,
+      "loss": 3.8655,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.804402032806089e-05,
+      "loss": 3.8753,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.803563438055037e-05,
+      "loss": 3.8867,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.802724843303985e-05,
+      "loss": 3.8727,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.801886248552933e-05,
+      "loss": 3.8805,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.801047653801881e-05,
+      "loss": 3.8681,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.800209059050829e-05,
+      "loss": 3.8726,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7993704642997765e-05,
+      "loss": 3.8716,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.798533507429098e-05,
+      "loss": 3.8758,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797694912678046e-05,
+      "loss": 3.8908,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796856317926994e-05,
+      "loss": 3.8812,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796019361056315e-05,
+      "loss": 3.8838,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.795180766305263e-05,
+      "loss": 3.874,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.794342171554211e-05,
+      "loss": 3.8736,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.793503576803159e-05,
+      "loss": 3.8761,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.792664982052107e-05,
+      "loss": 3.8833,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7918280251814282e-05,
+      "loss": 3.8829,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7909894304303762e-05,
+      "loss": 3.8815,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.790150835679324e-05,
+      "loss": 3.8802,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.789312240928272e-05,
+      "loss": 3.8734,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.78847364617722e-05,
+      "loss": 3.8825,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7876366893065414e-05,
+      "loss": 3.8756,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.786798094555489e-05,
+      "loss": 3.8699,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.785959499804437e-05,
+      "loss": 3.8819,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.785120905053385e-05,
+      "loss": 3.8722,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.784282310302333e-05,
+      "loss": 3.8743,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7834453534316543e-05,
+      "loss": 3.8711,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7826067586806023e-05,
+      "loss": 3.871,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7817681639295503e-05,
+      "loss": 3.8783,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7809295691784983e-05,
+      "loss": 3.8729,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7800909744274463e-05,
+      "loss": 3.8861,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7792523796763943e-05,
+      "loss": 3.8883,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.778413784925342e-05,
+      "loss": 3.8903,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7775768280546636e-05,
+      "loss": 3.8786,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7767382333036115e-05,
+      "loss": 3.8767,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7758996385525595e-05,
+      "loss": 3.8688,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7750610438015072e-05,
+      "loss": 3.8857,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7742240869308284e-05,
+      "loss": 3.8823,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7733854921797764e-05,
+      "loss": 3.8655,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7725468974287244e-05,
+      "loss": 3.8767,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7717083026776724e-05,
+      "loss": 3.8782,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7708713458069937e-05,
+      "loss": 3.8807,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7700327510559417e-05,
+      "loss": 3.8756,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7691941563048893e-05,
+      "loss": 3.8781,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7683555615538373e-05,
+      "loss": 3.8721,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7675169668027857e-05,
+      "loss": 3.8883,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7666783720517337e-05,
+      "loss": 3.8801,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7658397773006817e-05,
+      "loss": 3.8892,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7650011825496297e-05,
+      "loss": 3.8714,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7641625877985773e-05,
+      "loss": 3.8828,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7633239930475253e-05,
+      "loss": 3.8883,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7624870361768466e-05,
+      "loss": 3.8743,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7616484414257945e-05,
+      "loss": 3.8797,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7608098466747425e-05,
+      "loss": 3.878,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7599712519236905e-05,
+      "loss": 3.8833,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7591326571726385e-05,
+      "loss": 3.89,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7582957003019598e-05,
+      "loss": 3.8773,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7574571055509074e-05,
+      "loss": 3.8807,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7566185107998558e-05,
+      "loss": 3.8844,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7557799160488038e-05,
+      "loss": 3.8782,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7549413212977518e-05,
+      "loss": 3.8753,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7541043644270727e-05,
+      "loss": 3.8723,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7532657696760207e-05,
+      "loss": 3.8746,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7524271749249687e-05,
+      "loss": 3.8756,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7515885801739167e-05,
+      "loss": 3.881,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.987410545349121,
+      "eval_runtime": 296.8512,
+      "eval_samples_per_second": 1285.462,
+      "eval_steps_per_second": 40.172,
+      "step": 1984320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.750749985422865e-05,
+      "loss": 3.8831,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749911390671813e-05,
+      "loss": 3.8737,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7490727959207606e-05,
+      "loss": 3.8767,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7482342011697086e-05,
+      "loss": 3.8841,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7473956064186566e-05,
+      "loss": 3.8916,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7465570116676046e-05,
+      "loss": 3.8806,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7457184169165526e-05,
+      "loss": 3.8852,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7448798221655006e-05,
+      "loss": 3.8737,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.744041227414449e-05,
+      "loss": 3.8745,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7432026326633966e-05,
+      "loss": 3.8789,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7423640379123446e-05,
+      "loss": 3.8839,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7415254431612926e-05,
+      "loss": 3.8767,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7406884862906135e-05,
+      "loss": 3.8861,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7398498915395615e-05,
+      "loss": 3.8711,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.73901129678851e-05,
+      "loss": 3.8746,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.738172702037458e-05,
+      "loss": 3.8572,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.737334107286406e-05,
+      "loss": 3.8817,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.736495512535354e-05,
+      "loss": 3.8814,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7356569177843018e-05,
+      "loss": 3.8713,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7348183230332495e-05,
+      "loss": 3.8847,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7339797282821975e-05,
+      "loss": 3.8861,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7331411335311458e-05,
+      "loss": 3.8874,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.732304176660467e-05,
+      "loss": 3.8769,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7314655819094147e-05,
+      "loss": 3.8811,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7306269871583627e-05,
+      "loss": 3.8833,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7297883924073107e-05,
+      "loss": 3.8848,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7289497976562587e-05,
+      "loss": 3.8694,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.72811284078558e-05,
+      "loss": 3.8803,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.727274246034528e-05,
+      "loss": 3.8718,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.726435651283476e-05,
+      "loss": 3.8725,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.725597056532424e-05,
+      "loss": 3.879,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.724758461781372e-05,
+      "loss": 3.8813,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.72391986703032e-05,
+      "loss": 3.8798,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7230812722792676e-05,
+      "loss": 3.8876,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.722242677528216e-05,
+      "loss": 3.8751,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.72140735853791e-05,
+      "loss": 3.8836,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720568763786858e-05,
+      "loss": 3.8879,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719730169035806e-05,
+      "loss": 3.863,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718891574284754e-05,
+      "loss": 3.8692,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718052979533702e-05,
+      "loss": 3.8717,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7172160226630233e-05,
+      "loss": 3.8623,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7163774279119713e-05,
+      "loss": 3.8755,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7155388331609193e-05,
+      "loss": 3.8845,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7147002384098673e-05,
+      "loss": 3.869,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.713861643658815e-05,
+      "loss": 3.877,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.713023048907763e-05,
+      "loss": 3.8755,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7121860920370845e-05,
+      "loss": 3.8787,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7113491351664055e-05,
+      "loss": 3.8653,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7105105404153535e-05,
+      "loss": 3.8576,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7096719456643014e-05,
+      "loss": 3.8837,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7088333509132494e-05,
+      "loss": 3.8812,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7079947561621974e-05,
+      "loss": 3.8918,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7071561614111454e-05,
+      "loss": 3.8712,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7063175666600934e-05,
+      "loss": 3.8722,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7054789719090414e-05,
+      "loss": 3.8753,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7046403771579894e-05,
+      "loss": 3.8701,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7038017824069374e-05,
+      "loss": 3.8875,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7029648255362583e-05,
+      "loss": 3.8566,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7021262307852067e-05,
+      "loss": 3.9006,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7012876360341547e-05,
+      "loss": 3.8798,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7004490412831027e-05,
+      "loss": 3.8574,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6996104465320503e-05,
+      "loss": 3.8821,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6987718517809983e-05,
+      "loss": 3.8705,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6979332570299463e-05,
+      "loss": 3.8578,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6970946622788943e-05,
+      "loss": 3.8804,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6962577054082155e-05,
+      "loss": 3.8822,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6954191106571635e-05,
+      "loss": 3.8671,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6945821537864848e-05,
+      "loss": 3.8788,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6937435590354328e-05,
+      "loss": 3.8541,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6929049642843804e-05,
+      "loss": 3.8686,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.692068007413702e-05,
+      "loss": 3.8722,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.69122941266265e-05,
+      "loss": 3.8743,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6903908179115977e-05,
+      "loss": 3.8764,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6895522231605457e-05,
+      "loss": 3.8712,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6887136284094937e-05,
+      "loss": 3.8921,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6878750336584417e-05,
+      "loss": 3.8737,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6870364389073897e-05,
+      "loss": 3.8798,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6861978441563377e-05,
+      "loss": 3.8727,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.685359249405286e-05,
+      "loss": 3.8705,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6845206546542336e-05,
+      "loss": 3.8793,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6836820599031816e-05,
+      "loss": 3.8722,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.682845103032503e-05,
+      "loss": 3.872,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.682006508281451e-05,
+      "loss": 3.8849,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6811679135303985e-05,
+      "loss": 3.8693,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.68033095665972e-05,
+      "loss": 3.8601,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.679492361908668e-05,
+      "loss": 3.8708,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6786537671576158e-05,
+      "loss": 3.8843,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6778151724065638e-05,
+      "loss": 3.8646,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6769765776555118e-05,
+      "loss": 3.8796,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6761379829044598e-05,
+      "loss": 3.8681,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6752993881534078e-05,
+      "loss": 3.8632,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.674460793402356e-05,
+      "loss": 3.8654,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673623836531677e-05,
+      "loss": 3.8703,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.672785241780625e-05,
+      "loss": 3.8896,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671946647029573e-05,
+      "loss": 3.8744,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671108052278521e-05,
+      "loss": 3.8783,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.670269457527469e-05,
+      "loss": 3.8706,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.669430862776417e-05,
+      "loss": 3.8668,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.668592268025365e-05,
+      "loss": 3.8696,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.667753673274313e-05,
+      "loss": 3.8781,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666916716403634e-05,
+      "loss": 3.8789,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666079759532955e-05,
+      "loss": 3.8773,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.665241164781903e-05,
+      "loss": 3.8773,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6644025700308515e-05,
+      "loss": 3.8656,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.663563975279799e-05,
+      "loss": 3.8797,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.662725380528747e-05,
+      "loss": 3.8715,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.661886785777695e-05,
+      "loss": 3.8662,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.661048191026643e-05,
+      "loss": 3.8725,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6602112341559644e-05,
+      "loss": 3.8708,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6593726394049124e-05,
+      "loss": 3.87,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6585340446538603e-05,
+      "loss": 3.8642,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6576954499028083e-05,
+      "loss": 3.8666,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6568568551517563e-05,
+      "loss": 3.8714,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6560182604007043e-05,
+      "loss": 3.8747,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6551813035300252e-05,
+      "loss": 3.8763,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6543427087789732e-05,
+      "loss": 3.8816,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6535041140279216e-05,
+      "loss": 3.8876,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6526655192768696e-05,
+      "loss": 3.8738,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6518269245258172e-05,
+      "loss": 3.8728,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6509883297747652e-05,
+      "loss": 3.8654,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6501497350237132e-05,
+      "loss": 3.8787,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6493127781530345e-05,
+      "loss": 3.8762,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6484758212823557e-05,
+      "loss": 3.8667,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6476372265313037e-05,
+      "loss": 3.8661,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6467986317802517e-05,
+      "loss": 3.8776,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6459600370291994e-05,
+      "loss": 3.8766,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6451214422781474e-05,
+      "loss": 3.8665,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6442828475270954e-05,
+      "loss": 3.8711,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6434442527760437e-05,
+      "loss": 3.8728,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6426056580249917e-05,
+      "loss": 3.8793,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6417670632739397e-05,
+      "loss": 3.8772,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6409301064032606e-05,
+      "loss": 3.884,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6400915116522086e-05,
+      "loss": 3.865,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6392529169011566e-05,
+      "loss": 3.8791,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6384143221501046e-05,
+      "loss": 3.8832,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6375757273990526e-05,
+      "loss": 3.8676,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6367387705283738e-05,
+      "loss": 3.8756,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6359001757773218e-05,
+      "loss": 3.877,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6350615810262698e-05,
+      "loss": 3.8835,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6342229862752175e-05,
+      "loss": 3.8839,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6333843915241655e-05,
+      "loss": 3.8718,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.632547434653487e-05,
+      "loss": 3.8764,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.631708839902435e-05,
+      "loss": 3.8802,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6308702451513827e-05,
+      "loss": 3.872,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.630033288280704e-05,
+      "loss": 3.8693,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.629194693529652e-05,
+      "loss": 3.8702,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6283560987786e-05,
+      "loss": 3.8654,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6275191419079212e-05,
+      "loss": 3.8744,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6266805471568692e-05,
+      "loss": 3.8755,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.98535418510437,
+      "eval_runtime": 298.1639,
+      "eval_samples_per_second": 1279.803,
+      "eval_steps_per_second": 39.995,
+      "step": 2060640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6258419524058172e-05,
+      "loss": 3.8688,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625003357654765e-05,
+      "loss": 3.8722,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.624164762903713e-05,
+      "loss": 3.8755,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.623326168152661e-05,
+      "loss": 3.8785,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.622487573401609e-05,
+      "loss": 3.8856,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.621648978650557e-05,
+      "loss": 3.8765,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.620810383899505e-05,
+      "loss": 3.8817,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.619971789148453e-05,
+      "loss": 3.8708,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6191331943974008e-05,
+      "loss": 3.8655,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6182945996463488e-05,
+      "loss": 3.8731,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6174560048952968e-05,
+      "loss": 3.8848,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6166174101442448e-05,
+      "loss": 3.8665,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.615780453273566e-05,
+      "loss": 3.8834,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614941858522514e-05,
+      "loss": 3.8649,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614103263771462e-05,
+      "loss": 3.8678,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.61326466902041e-05,
+      "loss": 3.8564,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.612426074269358e-05,
+      "loss": 3.8764,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.611587479518306e-05,
+      "loss": 3.8748,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.610748884767254e-05,
+      "loss": 3.869,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6099119278965753e-05,
+      "loss": 3.8729,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6090733331455233e-05,
+      "loss": 3.8863,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6082347383944713e-05,
+      "loss": 3.879,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.607396143643419e-05,
+      "loss": 3.8766,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.606557548892367e-05,
+      "loss": 3.8734,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.605718954141315e-05,
+      "loss": 3.8829,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6048803593902632e-05,
+      "loss": 3.8754,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6040417646392112e-05,
+      "loss": 3.8678,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6032031698881592e-05,
+      "loss": 3.8741,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.602364575137107e-05,
+      "loss": 3.8685,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.601525980386055e-05,
+      "loss": 3.8649,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.600689023515376e-05,
+      "loss": 3.8791,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.599850428764324e-05,
+      "loss": 3.8717,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.599011834013272e-05,
+      "loss": 3.8811,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.59817323926222e-05,
+      "loss": 3.8831,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.597334644511168e-05,
+      "loss": 3.8707,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5964976876404894e-05,
+      "loss": 3.8789,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.595659092889437e-05,
+      "loss": 3.8788,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5948204981383854e-05,
+      "loss": 3.8642,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5939819033873333e-05,
+      "loss": 3.8678,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5931433086362813e-05,
+      "loss": 3.862,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5923063517656023e-05,
+      "loss": 3.8604,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5914677570145502e-05,
+      "loss": 3.8721,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5906291622634982e-05,
+      "loss": 3.8807,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5897905675124462e-05,
+      "loss": 3.8652,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5889519727613942e-05,
+      "loss": 3.8693,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5881150158907155e-05,
+      "loss": 3.8739,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5872764211396635e-05,
+      "loss": 3.8707,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5864378263886115e-05,
+      "loss": 3.8633,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5855992316375595e-05,
+      "loss": 3.8539,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.584760636886507e-05,
+      "loss": 3.8767,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5839220421354555e-05,
+      "loss": 3.882,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5830834473844035e-05,
+      "loss": 3.8835,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5822448526333515e-05,
+      "loss": 3.8683,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5814078957626724e-05,
+      "loss": 3.868,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5805709388919936e-05,
+      "loss": 3.8702,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.579733982021315e-05,
+      "loss": 3.8623,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.578895387270263e-05,
+      "loss": 3.8823,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.578056792519211e-05,
+      "loss": 3.8549,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.577218197768159e-05,
+      "loss": 3.8925,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.576379603017107e-05,
+      "loss": 3.875,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5755410082660545e-05,
+      "loss": 3.8561,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5747024135150025e-05,
+      "loss": 3.8744,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5738638187639508e-05,
+      "loss": 3.8674,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5730252240128988e-05,
+      "loss": 3.8531,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5721866292618468e-05,
+      "loss": 3.873,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5713480345107948e-05,
+      "loss": 3.8789,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5705094397597428e-05,
+      "loss": 3.8639,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5696724828890637e-05,
+      "loss": 3.8746,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.568835526018385e-05,
+      "loss": 3.8501,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.567996931267333e-05,
+      "loss": 3.8651,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5671599743966542e-05,
+      "loss": 3.8662,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.566321379645602e-05,
+      "loss": 3.8685,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.56548278489455e-05,
+      "loss": 3.8663,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.564644190143498e-05,
+      "loss": 3.8692,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5638055953924462e-05,
+      "loss": 3.8857,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5629670006413942e-05,
+      "loss": 3.8699,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5621284058903422e-05,
+      "loss": 3.8729,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.561291449019663e-05,
+      "loss": 3.8682,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.560452854268611e-05,
+      "loss": 3.8662,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.559614259517559e-05,
+      "loss": 3.8769,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.558775664766507e-05,
+      "loss": 3.8676,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.557937070015455e-05,
+      "loss": 3.8665,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.557098475264403e-05,
+      "loss": 3.8789,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.556259880513351e-05,
+      "loss": 3.8687,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.555421285762299e-05,
+      "loss": 3.851,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.554582691011247e-05,
+      "loss": 3.868,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.553745734140568e-05,
+      "loss": 3.8761,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5529071393895163e-05,
+      "loss": 3.8628,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5520685446384643e-05,
+      "loss": 3.8722,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5512299498874123e-05,
+      "loss": 3.8673,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5503913551363603e-05,
+      "loss": 3.8573,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5495527603853083e-05,
+      "loss": 3.8655,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548714165634256e-05,
+      "loss": 3.8627,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.547875570883204e-05,
+      "loss": 3.8791,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5470386140125255e-05,
+      "loss": 3.8722,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5462000192614735e-05,
+      "loss": 3.8763,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5453614245104212e-05,
+      "loss": 3.8632,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5445228297593692e-05,
+      "loss": 3.8641,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5436842350083172e-05,
+      "loss": 3.8684,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.542845640257265e-05,
+      "loss": 3.8755,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.542007045506213e-05,
+      "loss": 3.8741,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.541168450755161e-05,
+      "loss": 3.8709,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5403331317648557e-05,
+      "loss": 3.8708,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5394945370138033e-05,
+      "loss": 3.8625,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5386559422627513e-05,
+      "loss": 3.8754,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5378173475116993e-05,
+      "loss": 3.8646,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5369787527606473e-05,
+      "loss": 3.8677,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5361401580095956e-05,
+      "loss": 3.8668,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5353032011389165e-05,
+      "loss": 3.865,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5344646063878645e-05,
+      "loss": 3.8687,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5336260116368125e-05,
+      "loss": 3.8565,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5327874168857605e-05,
+      "loss": 3.8596,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5319488221347085e-05,
+      "loss": 3.8713,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5311102273836565e-05,
+      "loss": 3.8696,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5302732705129778e-05,
+      "loss": 3.8691,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5294346757619258e-05,
+      "loss": 3.8772,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5285960810108738e-05,
+      "loss": 3.8854,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5277574862598214e-05,
+      "loss": 3.866,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5269188915087694e-05,
+      "loss": 3.8634,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526081934638091e-05,
+      "loss": 3.8633,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5252433398870387e-05,
+      "loss": 3.8728,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5244047451359867e-05,
+      "loss": 3.8716,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5235661503849347e-05,
+      "loss": 3.8641,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5227275556338826e-05,
+      "loss": 3.8635,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5218905987632037e-05,
+      "loss": 3.8745,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5210520040121517e-05,
+      "loss": 3.8672,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5202134092610997e-05,
+      "loss": 3.8674,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5193748145100479e-05,
+      "loss": 3.8719,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5185378576393688e-05,
+      "loss": 3.869,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.517699262888317e-05,
+      "loss": 3.8672,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.516860668137265e-05,
+      "loss": 3.8753,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.516022073386213e-05,
+      "loss": 3.8766,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.515183478635161e-05,
+      "loss": 3.8636,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.514344883884109e-05,
+      "loss": 3.8758,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5135062891330568e-05,
+      "loss": 3.8769,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5126676943820048e-05,
+      "loss": 3.8665,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5118290996309528e-05,
+      "loss": 3.8669,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5109921427602742e-05,
+      "loss": 3.8699,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5101535480092218e-05,
+      "loss": 3.8803,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5093149532581698e-05,
+      "loss": 3.8787,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.508476358507118e-05,
+      "loss": 3.8708,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.507637763756066e-05,
+      "loss": 3.87,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.506800806885387e-05,
+      "loss": 3.8745,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.505962212134335e-05,
+      "loss": 3.8723,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.505123617383283e-05,
+      "loss": 3.8613,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.504285022632231e-05,
+      "loss": 3.8697,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.503446427881179e-05,
+      "loss": 3.8572,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5026094710105001e-05,
+      "loss": 3.8711,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5017708762594481e-05,
+      "loss": 3.8688,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.984039306640625,
+      "eval_runtime": 306.6552,
+      "eval_samples_per_second": 1244.365,
+      "eval_steps_per_second": 38.887,
+      "step": 2136960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5009322815083963e-05,
+      "loss": 3.8757,
+      "step": 2137088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5000936867573443e-05,
+      "loss": 3.8628,
+      "step": 2137600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4992567298866652e-05,
+      "loss": 3.8688,
+      "step": 2138112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4984181351356134e-05,
+      "loss": 3.8791,
+      "step": 2138624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4975795403845614e-05,
+      "loss": 3.8753,
+      "step": 2139136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4967409456335093e-05,
+      "loss": 3.8777,
+      "step": 2139648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4959039887628304e-05,
+      "loss": 3.8745,
+      "step": 2140160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4950653940117784e-05,
+      "loss": 3.8653,
+      "step": 2140672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4942267992607264e-05,
+      "loss": 3.8605,
+      "step": 2141184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4933882045096744e-05,
+      "loss": 3.8747,
+      "step": 2141696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4925496097586222e-05,
+      "loss": 3.8701,
+      "step": 2142208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4917110150075702e-05,
+      "loss": 3.8681,
+      "step": 2142720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4908740581368917e-05,
+      "loss": 3.8792,
+      "step": 2143232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4900354633858396e-05,
+      "loss": 3.8574,
+      "step": 2143744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4891968686347873e-05,
+      "loss": 3.8688,
+      "step": 2144256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4883582738837355e-05,
+      "loss": 3.8506,
+      "step": 2144768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4875196791326835e-05,
+      "loss": 3.8721,
+      "step": 2145280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4866810843816315e-05,
+      "loss": 3.867,
+      "step": 2145792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4858424896305795e-05,
+      "loss": 3.8674,
+      "step": 2146304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4850038948795275e-05,
+      "loss": 3.8713,
+      "step": 2146816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4841653001284756e-05,
+      "loss": 3.8789,
+      "step": 2147328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4833283432577965e-05,
+      "loss": 3.8749,
+      "step": 2147840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4824897485067445e-05,
+      "loss": 3.8703,
+      "step": 2148352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4816527916360656e-05,
+      "loss": 3.8687,
+      "step": 2148864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4808141968850136e-05,
+      "loss": 3.882,
+      "step": 2149376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4799756021339618e-05,
+      "loss": 3.8708,
+      "step": 2149888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4791370073829098e-05,
+      "loss": 3.8605,
+      "step": 2150400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4783000505122307e-05,
+      "loss": 3.8706,
+      "step": 2150912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4774614557611788e-05,
+      "loss": 3.8583,
+      "step": 2151424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4766244988904997e-05,
+      "loss": 3.867,
+      "step": 2151936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4757859041394479e-05,
+      "loss": 3.8754,
+      "step": 2152448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4749473093883959e-05,
+      "loss": 3.8666,
+      "step": 2152960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4741087146373439e-05,
+      "loss": 3.8712,
+      "step": 2153472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4732701198862919e-05,
+      "loss": 3.8771,
+      "step": 2153984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4724315251352399e-05,
+      "loss": 3.8672,
+      "step": 2154496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.471594568264561e-05,
+      "loss": 3.8729,
+      "step": 2155008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.470755973513509e-05,
+      "loss": 3.8756,
+      "step": 2155520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4699173787624571e-05,
+      "loss": 3.858,
+      "step": 2156032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4690787840114051e-05,
+      "loss": 3.8664,
+      "step": 2156544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4682401892603528e-05,
+      "loss": 3.8573,
+      "step": 2157056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.467401594509301e-05,
+      "loss": 3.8569,
+      "step": 2157568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.466562999758249e-05,
+      "loss": 3.8675,
+      "step": 2158080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.465724405007197e-05,
+      "loss": 3.874,
+      "step": 2158592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.464885810256145e-05,
+      "loss": 3.8585,
+      "step": 2159104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.464047215505093e-05,
+      "loss": 3.8666,
+      "step": 2159616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4632086207540408e-05,
+      "loss": 3.8698,
+      "step": 2160128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.462371663883362e-05,
+      "loss": 3.8701,
+      "step": 2160640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4615330691323102e-05,
+      "loss": 3.8606,
+      "step": 2161152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4606944743812582e-05,
+      "loss": 3.8476,
+      "step": 2161664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4598558796302058e-05,
+      "loss": 3.8679,
+      "step": 2162176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4590189227595272e-05,
+      "loss": 3.8823,
+      "step": 2162688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4581803280084752e-05,
+      "loss": 3.8784,
+      "step": 2163200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4573417332574232e-05,
+      "loss": 3.8597,
+      "step": 2163712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4565047763867443e-05,
+      "loss": 3.8646,
+      "step": 2164224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4556661816356923e-05,
+      "loss": 3.8699,
+      "step": 2164736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4548275868846403e-05,
+      "loss": 3.8533,
+      "step": 2165248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4539889921335881e-05,
+      "loss": 3.8849,
+      "step": 2165760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4531503973825361e-05,
+      "loss": 3.8486,
+      "step": 2166272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4523118026314841e-05,
+      "loss": 3.8853,
+      "step": 2166784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4514732078804321e-05,
+      "loss": 3.8702,
+      "step": 2167296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4506346131293803e-05,
+      "loss": 3.8544,
+      "step": 2167808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4497960183783283e-05,
+      "loss": 3.8673,
+      "step": 2168320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4489574236272763e-05,
+      "loss": 3.865,
+      "step": 2168832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4481188288762241e-05,
+      "loss": 3.8471,
+      "step": 2169344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4472802341251721e-05,
+      "loss": 3.8688,
+      "step": 2169856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4464432772544933e-05,
+      "loss": 3.8757,
+      "step": 2170368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4456046825034413e-05,
+      "loss": 3.8619,
+      "step": 2170880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4447660877523892e-05,
+      "loss": 3.8693,
+      "step": 2171392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4439274930013372e-05,
+      "loss": 3.8452,
+      "step": 2171904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4430921740110315e-05,
+      "loss": 3.8592,
+      "step": 2172416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4422535792599795e-05,
+      "loss": 3.8611,
+      "step": 2172928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4414149845089275e-05,
+      "loss": 3.8683,
+      "step": 2173440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4405763897578756e-05,
+      "loss": 3.8615,
+      "step": 2173952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4397377950068236e-05,
+      "loss": 3.8666,
+      "step": 2174464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4388992002557713e-05,
+      "loss": 3.8799,
+      "step": 2174976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4380622433850927e-05,
+      "loss": 3.8669,
+      "step": 2175488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4372236486340407e-05,
+      "loss": 3.868,
+      "step": 2176000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4363850538829887e-05,
+      "loss": 3.8663,
+      "step": 2176512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4355464591319365e-05,
+      "loss": 3.8613,
+      "step": 2177024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4347078643808845e-05,
+      "loss": 3.8713,
+      "step": 2177536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4338709075102058e-05,
+      "loss": 3.8624,
+      "step": 2178048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4330323127591536e-05,
+      "loss": 3.8667,
+      "step": 2178560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4321937180081016e-05,
+      "loss": 3.8705,
+      "step": 2179072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4313551232570496e-05,
+      "loss": 3.8682,
+      "step": 2179584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4305165285059976e-05,
+      "loss": 3.8449,
+      "step": 2180096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4296779337549458e-05,
+      "loss": 3.8674,
+      "step": 2180608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4288393390038937e-05,
+      "loss": 3.8668,
+      "step": 2181120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4280007442528417e-05,
+      "loss": 3.8585,
+      "step": 2181632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.427165425262536e-05,
+      "loss": 3.8674,
+      "step": 2182144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4263268305114837e-05,
+      "loss": 3.8601,
+      "step": 2182656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4254882357604319e-05,
+      "loss": 3.8563,
+      "step": 2183168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4246496410093799e-05,
+      "loss": 3.8586,
+      "step": 2183680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4238110462583279e-05,
+      "loss": 3.8629,
+      "step": 2184192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4229724515072759e-05,
+      "loss": 3.8727,
+      "step": 2184704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4221338567562239e-05,
+      "loss": 3.8685,
+      "step": 2185216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4212952620051717e-05,
+      "loss": 3.8708,
+      "step": 2185728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4204566672541197e-05,
+      "loss": 3.8612,
+      "step": 2186240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4196197103834411e-05,
+      "loss": 3.8576,
+      "step": 2186752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4187811156323891e-05,
+      "loss": 3.8622,
+      "step": 2187264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4179425208813368e-05,
+      "loss": 3.8698,
+      "step": 2187776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.417103926130285e-05,
+      "loss": 3.8697,
+      "step": 2188288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4162669692596062e-05,
+      "loss": 3.8657,
+      "step": 2188800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4154300123889273e-05,
+      "loss": 3.8669,
+      "step": 2189312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4145914176378753e-05,
+      "loss": 3.8585,
+      "step": 2189824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4137528228868233e-05,
+      "loss": 3.8691,
+      "step": 2190336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4129142281357713e-05,
+      "loss": 3.8638,
+      "step": 2190848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.412075633384719e-05,
+      "loss": 3.8625,
+      "step": 2191360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4112386765140403e-05,
+      "loss": 3.8612,
+      "step": 2191872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4104000817629883e-05,
+      "loss": 3.8618,
+      "step": 2192384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4095614870119365e-05,
+      "loss": 3.8603,
+      "step": 2192896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4087228922608841e-05,
+      "loss": 3.8573,
+      "step": 2193408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4078842975098321e-05,
+      "loss": 3.8506,
+      "step": 2193920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4070473406391536e-05,
+      "loss": 3.8657,
+      "step": 2194432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4062087458881012e-05,
+      "loss": 3.8712,
+      "step": 2194944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4053701511370492e-05,
+      "loss": 3.8608,
+      "step": 2195456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4045315563859974e-05,
+      "loss": 3.8717,
+      "step": 2195968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4036945995153186e-05,
+      "loss": 3.8825,
+      "step": 2196480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4028560047642664e-05,
+      "loss": 3.866,
+      "step": 2196992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4020174100132144e-05,
+      "loss": 3.8619,
+      "step": 2197504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4011788152621624e-05,
+      "loss": 3.8613,
+      "step": 2198016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4003402205111104e-05,
+      "loss": 3.8662,
+      "step": 2198528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3995016257600584e-05,
+      "loss": 3.8646,
+      "step": 2199040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3986630310090066e-05,
+      "loss": 3.8611,
+      "step": 2199552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3978244362579546e-05,
+      "loss": 3.8615,
+      "step": 2200064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3969891172676486e-05,
+      "loss": 3.8683,
+      "step": 2200576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3961505225165966e-05,
+      "loss": 3.8624,
+      "step": 2201088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3953119277655446e-05,
+      "loss": 3.868,
+      "step": 2201600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.394474970894866e-05,
+      "loss": 3.8653,
+      "step": 2202112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3936363761438137e-05,
+      "loss": 3.8637,
+      "step": 2202624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3927977813927618e-05,
+      "loss": 3.8609,
+      "step": 2203136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3919591866417098e-05,
+      "loss": 3.8765,
+      "step": 2203648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3911205918906578e-05,
+      "loss": 3.8704,
+      "step": 2204160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3902819971396058e-05,
+      "loss": 3.8607,
+      "step": 2204672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3894434023885538e-05,
+      "loss": 3.8668,
+      "step": 2205184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.388604807637502e-05,
+      "loss": 3.8746,
+      "step": 2205696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3877662128864496e-05,
+      "loss": 3.86,
+      "step": 2206208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3869276181353976e-05,
+      "loss": 3.8672,
+      "step": 2206720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.386090661264719e-05,
+      "loss": 3.8676,
+      "step": 2207232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3852520665136667e-05,
+      "loss": 3.8771,
+      "step": 2207744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3844134717626147e-05,
+      "loss": 3.8717,
+      "step": 2208256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3835748770115629e-05,
+      "loss": 3.8709,
+      "step": 2208768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3827362822605108e-05,
+      "loss": 3.8636,
+      "step": 2209280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3818976875094588e-05,
+      "loss": 3.8753,
+      "step": 2209792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3810590927584068e-05,
+      "loss": 3.8649,
+      "step": 2210304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.380220498007355e-05,
+      "loss": 3.8575,
+      "step": 2210816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3793819032563027e-05,
+      "loss": 3.8647,
+      "step": 2211328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3785449463856239e-05,
+      "loss": 3.8537,
+      "step": 2211840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.377706351634572e-05,
+      "loss": 3.8682,
+      "step": 2212352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.37686775688352e-05,
+      "loss": 3.8632,
+      "step": 2212864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.982963800430298,
+      "eval_runtime": 300.261,
+      "eval_samples_per_second": 1270.865,
+      "eval_steps_per_second": 39.715,
+      "step": 2213280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3760308000128411e-05,
+      "loss": 3.8715,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3751922052617891e-05,
+      "loss": 3.8612,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3743536105107371e-05,
+      "loss": 3.8654,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.373515015759685e-05,
+      "loss": 3.8754,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.372676421008633e-05,
+      "loss": 3.8701,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.371837826257581e-05,
+      "loss": 3.8789,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.370999231506529e-05,
+      "loss": 3.8678,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.370160636755477e-05,
+      "loss": 3.8622,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3693220420044251e-05,
+      "loss": 3.8516,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3684834472533731e-05,
+      "loss": 3.871,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3676448525023208e-05,
+      "loss": 3.8684,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.366806257751269e-05,
+      "loss": 3.8643,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3659693008805902e-05,
+      "loss": 3.8799,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3651307061295382e-05,
+      "loss": 3.85,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.364292111378486e-05,
+      "loss": 3.8606,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.363453516627434e-05,
+      "loss": 3.855,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.362614921876382e-05,
+      "loss": 3.8642,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.36177632712533e-05,
+      "loss": 3.8657,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.360939370254651e-05,
+      "loss": 3.8585,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.360100775503599e-05,
+      "loss": 3.8698,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.359262180752547e-05,
+      "loss": 3.8729,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3584235860014952e-05,
+      "loss": 3.8717,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3575849912504432e-05,
+      "loss": 3.8678,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3567463964993912e-05,
+      "loss": 3.8647,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.355907801748339e-05,
+      "loss": 3.8783,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.355069206997287e-05,
+      "loss": 3.8635,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.354230612246235e-05,
+      "loss": 3.86,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3533936553755561e-05,
+      "loss": 3.8662,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3525550606245041e-05,
+      "loss": 3.8552,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3517164658734521e-05,
+      "loss": 3.8612,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3508778711224001e-05,
+      "loss": 3.8672,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3500392763713483e-05,
+      "loss": 3.8631,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3492023195006692e-05,
+      "loss": 3.8714,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3483637247496173e-05,
+      "loss": 3.8732,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3475251299985653e-05,
+      "loss": 3.8629,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3466881731278862e-05,
+      "loss": 3.871,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3458495783768344e-05,
+      "loss": 3.8743,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3450109836257824e-05,
+      "loss": 3.8529,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3441723888747304e-05,
+      "loss": 3.8608,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3433337941236784e-05,
+      "loss": 3.8501,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3424951993726264e-05,
+      "loss": 3.8559,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3416566046215742e-05,
+      "loss": 3.8601,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3408180098705222e-05,
+      "loss": 3.8752,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3399794151194702e-05,
+      "loss": 3.8534,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3391440961291645e-05,
+      "loss": 3.8614,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3383055013781125e-05,
+      "loss": 3.8637,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3374669066270607e-05,
+      "loss": 3.8641,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3366283118760087e-05,
+      "loss": 3.8564,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3357897171249567e-05,
+      "loss": 3.848,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3349511223739045e-05,
+      "loss": 3.865,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3341125276228525e-05,
+      "loss": 3.8739,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3332755707521738e-05,
+      "loss": 3.8708,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3324369760011216e-05,
+      "loss": 3.8593,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3315983812500696e-05,
+      "loss": 3.8553,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3307597864990176e-05,
+      "loss": 3.8691,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3299211917479656e-05,
+      "loss": 3.849,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3290825969969137e-05,
+      "loss": 3.8794,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3282440022458617e-05,
+      "loss": 3.8457,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3274070453751828e-05,
+      "loss": 3.8812,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3265684506241308e-05,
+      "loss": 3.8665,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3257298558730788e-05,
+      "loss": 3.8461,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3248912611220268e-05,
+      "loss": 3.8614,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3240526663709748e-05,
+      "loss": 3.8637,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3232140716199226e-05,
+      "loss": 3.8475,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3223754768688706e-05,
+      "loss": 3.856,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215368821178186e-05,
+      "loss": 3.8766,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3206999252471397e-05,
+      "loss": 3.8604,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3198613304960877e-05,
+      "loss": 3.861,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3190227357450358e-05,
+      "loss": 3.8434,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3181841409939838e-05,
+      "loss": 3.8599,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3173455462429318e-05,
+      "loss": 3.8544,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3165069514918798e-05,
+      "loss": 3.8639,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3156699946212009e-05,
+      "loss": 3.8583,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3148313998701489e-05,
+      "loss": 3.8623,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3139928051190969e-05,
+      "loss": 3.8724,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3131542103680449e-05,
+      "loss": 3.862,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.312315615616993e-05,
+      "loss": 3.862,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3114770208659407e-05,
+      "loss": 3.8638,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3106400639952621e-05,
+      "loss": 3.8615,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3098014692442101e-05,
+      "loss": 3.8657,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.308964512373531e-05,
+      "loss": 3.8569,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3081259176224792e-05,
+      "loss": 3.8605,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3072873228714272e-05,
+      "loss": 3.8689,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3064487281203752e-05,
+      "loss": 3.8628,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.305610133369323e-05,
+      "loss": 3.8439,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.304771538618271e-05,
+      "loss": 3.8622,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.303932943867219e-05,
+      "loss": 3.8597,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.303094349116167e-05,
+      "loss": 3.861,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3022573922454881e-05,
+      "loss": 3.8585,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3014187974944361e-05,
+      "loss": 3.8634,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3005802027433841e-05,
+      "loss": 3.849,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2997416079923323e-05,
+      "loss": 3.8548,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2989030132412802e-05,
+      "loss": 3.8568,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2980660563706013e-05,
+      "loss": 3.8714,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2972290994999226e-05,
+      "loss": 3.8629,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2963905047488702e-05,
+      "loss": 3.8703,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2955535478781916e-05,
+      "loss": 3.8544,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2947149531271396e-05,
+      "loss": 3.8565,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2938763583760875e-05,
+      "loss": 3.8599,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2930377636250355e-05,
+      "loss": 3.867,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2921991688739835e-05,
+      "loss": 3.864,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2913605741229315e-05,
+      "loss": 3.8635,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2905219793718795e-05,
+      "loss": 3.8639,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2896833846208276e-05,
+      "loss": 3.8521,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2888447898697756e-05,
+      "loss": 3.8713,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2880061951187233e-05,
+      "loss": 3.8556,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2871676003676714e-05,
+      "loss": 3.8591,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2863290056166194e-05,
+      "loss": 3.8581,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2854920487459407e-05,
+      "loss": 3.8567,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2846550918752618e-05,
+      "loss": 3.8564,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2838164971242098e-05,
+      "loss": 3.8606,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2829779023731577e-05,
+      "loss": 3.8463,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2821393076221056e-05,
+      "loss": 3.862,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2813007128710536e-05,
+      "loss": 3.8657,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2804621181200016e-05,
+      "loss": 3.8552,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2796235233689496e-05,
+      "loss": 3.8711,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2787849286178977e-05,
+      "loss": 3.8753,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2779479717472186e-05,
+      "loss": 3.8612,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2771093769961668e-05,
+      "loss": 3.8585,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2762707822451148e-05,
+      "loss": 3.8554,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2754321874940628e-05,
+      "loss": 3.8617,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2745935927430108e-05,
+      "loss": 3.8611,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2737549979919588e-05,
+      "loss": 3.8577,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2729164032409066e-05,
+      "loss": 3.8555,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2720778084898546e-05,
+      "loss": 3.8637,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2712392137388026e-05,
+      "loss": 3.8614,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2704006189877508e-05,
+      "loss": 3.8581,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2695620242366988e-05,
+      "loss": 3.8598,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2687250673660198e-05,
+      "loss": 3.8607,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2678864726149678e-05,
+      "loss": 3.8545,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2670478778639158e-05,
+      "loss": 3.8734,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2662092831128638e-05,
+      "loss": 3.8692,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2653723262421849e-05,
+      "loss": 3.8601,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2645337314911329e-05,
+      "loss": 3.8616,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2636951367400809e-05,
+      "loss": 3.8675,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2628565419890289e-05,
+      "loss": 3.8609,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.262017947237977e-05,
+      "loss": 3.8628,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2611793524869247e-05,
+      "loss": 3.8609,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2603407577358727e-05,
+      "loss": 3.8786,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2595021629848209e-05,
+      "loss": 3.861,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2586635682337689e-05,
+      "loss": 3.8687,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.25782661136309e-05,
+      "loss": 3.8608,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2569896544924112e-05,
+      "loss": 3.8737,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2561510597413592e-05,
+      "loss": 3.8599,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.255312464990307e-05,
+      "loss": 3.8525,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.254473870239255e-05,
+      "loss": 3.863,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.253635275488203e-05,
+      "loss": 3.8503,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.252796680737151e-05,
+      "loss": 3.8641,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2519580859860992e-05,
+      "loss": 3.8571,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9817779064178467,
+      "eval_runtime": 294.5022,
+      "eval_samples_per_second": 1295.715,
+      "eval_steps_per_second": 40.492,
+      "step": 2289600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2511194912350472e-05,
+      "loss": 3.8635,
+      "step": 2289664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2502808964839952e-05,
+      "loss": 3.8599,
+      "step": 2290176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.249442301732943e-05,
+      "loss": 3.8584,
+      "step": 2290688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.248603706981891e-05,
+      "loss": 3.8724,
+      "step": 2291200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.247766750111212e-05,
+      "loss": 3.8656,
+      "step": 2291712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.24692815536016e-05,
+      "loss": 3.8753,
+      "step": 2292224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2460895606091082e-05,
+      "loss": 3.8595,
+      "step": 2292736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.245250965858056e-05,
+      "loss": 3.8581,
+      "step": 2293248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2444156468677504e-05,
+      "loss": 3.8542,
+      "step": 2293760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2435770521166984e-05,
+      "loss": 3.8653,
+      "step": 2294272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2427384573656464e-05,
+      "loss": 3.8624,
+      "step": 2294784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2418998626145944e-05,
+      "loss": 3.8603,
+      "step": 2295296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2410645436242887e-05,
+      "loss": 3.877,
+      "step": 2295808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2402259488732367e-05,
+      "loss": 3.85,
+      "step": 2296320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2393873541221847e-05,
+      "loss": 3.8541,
+      "step": 2296832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2385487593711325e-05,
+      "loss": 3.8514,
+      "step": 2297344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2377101646200807e-05,
+      "loss": 3.8582,
+      "step": 2297856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2368715698690287e-05,
+      "loss": 3.8614,
+      "step": 2298368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2360329751179765e-05,
+      "loss": 3.8577,
+      "step": 2298880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2351943803669245e-05,
+      "loss": 3.8626,
+      "step": 2299392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2343557856158727e-05,
+      "loss": 3.8707,
+      "step": 2299904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2335171908648205e-05,
+      "loss": 3.863,
+      "step": 2300416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2326785961137685e-05,
+      "loss": 3.8657,
+      "step": 2300928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2318400013627165e-05,
+      "loss": 3.8577,
+      "step": 2301440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2310014066116646e-05,
+      "loss": 3.8748,
+      "step": 2301952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2301644497409856e-05,
+      "loss": 3.8608,
+      "step": 2302464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2293274928703068e-05,
+      "loss": 3.8598,
+      "step": 2302976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2284888981192548e-05,
+      "loss": 3.8572,
+      "step": 2303488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2276503033682028e-05,
+      "loss": 3.8547,
+      "step": 2304000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2268117086171508e-05,
+      "loss": 3.8597,
+      "step": 2304512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2259731138660988e-05,
+      "loss": 3.8611,
+      "step": 2305024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2251345191150468e-05,
+      "loss": 3.8609,
+      "step": 2305536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2242959243639946e-05,
+      "loss": 3.8667,
+      "step": 2306048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2234573296129428e-05,
+      "loss": 3.8706,
+      "step": 2306560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2226187348618908e-05,
+      "loss": 3.8585,
+      "step": 2307072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2217817779912118e-05,
+      "loss": 3.8688,
+      "step": 2307584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2209431832401598e-05,
+      "loss": 3.8649,
+      "step": 2308096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2201045884891078e-05,
+      "loss": 3.8518,
+      "step": 2308608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2192659937380558e-05,
+      "loss": 3.8577,
+      "step": 2309120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2184290368673769e-05,
+      "loss": 3.845,
+      "step": 2309632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2175904421163249e-05,
+      "loss": 3.8557,
+      "step": 2310144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2167518473652729e-05,
+      "loss": 3.8558,
+      "step": 2310656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2159132526142209e-05,
+      "loss": 3.8693,
+      "step": 2311168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2150746578631689e-05,
+      "loss": 3.8476,
+      "step": 2311680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.21423770099249e-05,
+      "loss": 3.8571,
+      "step": 2312192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2133991062414381e-05,
+      "loss": 3.8624,
+      "step": 2312704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.212560511490386e-05,
+      "loss": 3.8601,
+      "step": 2313216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.211721916739334e-05,
+      "loss": 3.8583,
+      "step": 2313728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2108849598686552e-05,
+      "loss": 3.8381,
+      "step": 2314240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2100463651176032e-05,
+      "loss": 3.8596,
+      "step": 2314752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.209207770366551e-05,
+      "loss": 3.8726,
+      "step": 2315264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2083691756154992e-05,
+      "loss": 3.8683,
+      "step": 2315776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2075305808644472e-05,
+      "loss": 3.8563,
+      "step": 2316288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.206691986113395e-05,
+      "loss": 3.8532,
+      "step": 2316800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.205853391362343e-05,
+      "loss": 3.8662,
+      "step": 2317312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2050164344916643e-05,
+      "loss": 3.8481,
+      "step": 2317824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2041778397406123e-05,
+      "loss": 3.8728,
+      "step": 2318336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2033408828699333e-05,
+      "loss": 3.842,
+      "step": 2318848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2025022881188813e-05,
+      "loss": 3.8793,
+      "step": 2319360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2016636933678293e-05,
+      "loss": 3.8653,
+      "step": 2319872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2008250986167773e-05,
+      "loss": 3.8429,
+      "step": 2320384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1999865038657253e-05,
+      "loss": 3.8552,
+      "step": 2320896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1991479091146733e-05,
+      "loss": 3.8656,
+      "step": 2321408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1983093143636213e-05,
+      "loss": 3.8411,
+      "step": 2321920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1974707196125693e-05,
+      "loss": 3.8478,
+      "step": 2322432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1966337627418904e-05,
+      "loss": 3.8727,
+      "step": 2322944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1957951679908384e-05,
+      "loss": 3.8571,
+      "step": 2323456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1949565732397865e-05,
+      "loss": 3.8555,
+      "step": 2323968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1941179784887344e-05,
+      "loss": 3.8383,
+      "step": 2324480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1932793837376824e-05,
+      "loss": 3.8549,
+      "step": 2324992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1924407889866304e-05,
+      "loss": 3.8522,
+      "step": 2325504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1916021942355784e-05,
+      "loss": 3.8635,
+      "step": 2326016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1907635994845264e-05,
+      "loss": 3.8513,
+      "step": 2326528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1899266426138474e-05,
+      "loss": 3.8622,
+      "step": 2327040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1890896857431687e-05,
+      "loss": 3.8659,
+      "step": 2327552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1882527288724898e-05,
+      "loss": 3.8616,
+      "step": 2328064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1874141341214378e-05,
+      "loss": 3.8562,
+      "step": 2328576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1865755393703858e-05,
+      "loss": 3.8572,
+      "step": 2329088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1857369446193337e-05,
+      "loss": 3.8608,
+      "step": 2329600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1848983498682817e-05,
+      "loss": 3.8643,
+      "step": 2330112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1840597551172297e-05,
+      "loss": 3.8529,
+      "step": 2330624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1832211603661777e-05,
+      "loss": 3.8564,
+      "step": 2331136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1823825656151256e-05,
+      "loss": 3.8611,
+      "step": 2331648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1815439708640737e-05,
+      "loss": 3.8621,
+      "step": 2332160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1807070139933948e-05,
+      "loss": 3.8412,
+      "step": 2332672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1798684192423428e-05,
+      "loss": 3.8573,
+      "step": 2333184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1790298244912908e-05,
+      "loss": 3.8532,
+      "step": 2333696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1781912297402388e-05,
+      "loss": 3.8578,
+      "step": 2334208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1773526349891868e-05,
+      "loss": 3.8589,
+      "step": 2334720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1765140402381348e-05,
+      "loss": 3.8586,
+      "step": 2335232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1756770833674559e-05,
+      "loss": 3.8482,
+      "step": 2335744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1748384886164039e-05,
+      "loss": 3.8483,
+      "step": 2336256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1739998938653519e-05,
+      "loss": 3.8515,
+      "step": 2336768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.173162936994673e-05,
+      "loss": 3.8648,
+      "step": 2337280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.172324342243621e-05,
+      "loss": 3.8597,
+      "step": 2337792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1714857474925691e-05,
+      "loss": 3.8629,
+      "step": 2338304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.170647152741517e-05,
+      "loss": 3.8542,
+      "step": 2338816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1698101958708382e-05,
+      "loss": 3.8552,
+      "step": 2339328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1689716011197862e-05,
+      "loss": 3.8496,
+      "step": 2339840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1681330063687342e-05,
+      "loss": 3.8644,
+      "step": 2340352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.167294411617682e-05,
+      "loss": 3.8579,
+      "step": 2340864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1664558168666301e-05,
+      "loss": 3.8597,
+      "step": 2341376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1656188599959512e-05,
+      "loss": 3.8613,
+      "step": 2341888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.164780265244899e-05,
+      "loss": 3.8495,
+      "step": 2342400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1639416704938472e-05,
+      "loss": 3.8697,
+      "step": 2342912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1631030757427952e-05,
+      "loss": 3.8483,
+      "step": 2343424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1622644809917432e-05,
+      "loss": 3.8539,
+      "step": 2343936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1614258862406912e-05,
+      "loss": 3.8593,
+      "step": 2344448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1605872914896392e-05,
+      "loss": 3.851,
+      "step": 2344960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1597486967385872e-05,
+      "loss": 3.8522,
+      "step": 2345472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1589117398679083e-05,
+      "loss": 3.8554,
+      "step": 2345984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1580731451168563e-05,
+      "loss": 3.8432,
+      "step": 2346496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1572345503658043e-05,
+      "loss": 3.8593,
+      "step": 2347008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1563959556147523e-05,
+      "loss": 3.8571,
+      "step": 2347520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1555573608637003e-05,
+      "loss": 3.8569,
+      "step": 2348032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1547187661126483e-05,
+      "loss": 3.8673,
+      "step": 2348544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1538818092419693e-05,
+      "loss": 3.8712,
+      "step": 2349056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1530432144909173e-05,
+      "loss": 3.8597,
+      "step": 2349568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1522046197398653e-05,
+      "loss": 3.8554,
+      "step": 2350080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1513660249888133e-05,
+      "loss": 3.8501,
+      "step": 2350592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1505274302377613e-05,
+      "loss": 3.8574,
+      "step": 2351104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1496888354867093e-05,
+      "loss": 3.853,
+      "step": 2351616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1488502407356573e-05,
+      "loss": 3.8574,
+      "step": 2352128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1480116459846053e-05,
+      "loss": 3.8476,
+      "step": 2352640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1471746891139264e-05,
+      "loss": 3.8636,
+      "step": 2353152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1463360943628744e-05,
+      "loss": 3.8557,
+      "step": 2353664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1454974996118224e-05,
+      "loss": 3.8588,
+      "step": 2354176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1446589048607705e-05,
+      "loss": 3.8578,
+      "step": 2354688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1438219479900914e-05,
+      "loss": 3.856,
+      "step": 2355200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1429833532390394e-05,
+      "loss": 3.8485,
+      "step": 2355712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1421447584879876e-05,
+      "loss": 3.8656,
+      "step": 2356224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1413061637369354e-05,
+      "loss": 3.8657,
+      "step": 2356736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1404675689858834e-05,
+      "loss": 3.8601,
+      "step": 2357248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1396289742348314e-05,
+      "loss": 3.8535,
+      "step": 2357760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1387920173641527e-05,
+      "loss": 3.8614,
+      "step": 2358272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1379534226131005e-05,
+      "loss": 3.8582,
+      "step": 2358784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1371148278620487e-05,
+      "loss": 3.8593,
+      "step": 2359296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1362762331109967e-05,
+      "loss": 3.8627,
+      "step": 2359808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1354376383599445e-05,
+      "loss": 3.8702,
+      "step": 2360320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1346006814892657e-05,
+      "loss": 3.8587,
+      "step": 2360832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1337620867382137e-05,
+      "loss": 3.8668,
+      "step": 2361344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1329234919871617e-05,
+      "loss": 3.8571,
+      "step": 2361856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1320848972361097e-05,
+      "loss": 3.8635,
+      "step": 2362368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1312479403654308e-05,
+      "loss": 3.8537,
+      "step": 2362880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1304093456143788e-05,
+      "loss": 3.8544,
+      "step": 2363392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1295707508633268e-05,
+      "loss": 3.8611,
+      "step": 2363904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1287321561122748e-05,
+      "loss": 3.8438,
+      "step": 2364416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1278935613612228e-05,
+      "loss": 3.8571,
+      "step": 2364928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1270549666101708e-05,
+      "loss": 3.8571,
+      "step": 2365440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.980487108230591,
+      "eval_runtime": 298.8864,
+      "eval_samples_per_second": 1276.709,
+      "eval_steps_per_second": 39.898,
+      "step": 2365920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 9.590669932458861e+17,
+  "trial_name": null,
+  "trial_params": null
+}