diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 4.111427307128906,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/lstm/2/checkpoints/checkpoint-686880",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "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": 10.8206,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.554,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0624,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 7.0013,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9543,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9143,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7423,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6408,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5414,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4563,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3913,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3289,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.261,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989101543997071e-05,
+      "loss": 6.1967,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988262949246018e-05,
+      "loss": 6.1493,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987424354494966e-05,
+      "loss": 6.0967,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986585759743914e-05,
+      "loss": 6.0539,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985747164992862e-05,
+      "loss": 6.0146,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.9802,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.9423,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.9104,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8715,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8492,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.8079,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7899,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7587,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.7358,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.7155,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6849,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6593,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 5.6479,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.6353,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.6146,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5901,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5845,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5581,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 5.5454,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 5.5124,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 5.5149,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.493,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4842,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4728,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 5.4662,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 5.4354,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.4295,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.4243,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.409,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.4109,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3885,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3679,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3762,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3544,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3493,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.3211,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.3322,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.3092,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.3357,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2975,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2987,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2902,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2708,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2577,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2586,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2407,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.2347,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 5.2516,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 5.2333,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 5.218,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 5.1964,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.1919,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.2002,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1976,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1903,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1874,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1781,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.171,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1569,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1484,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1412,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 5.1299,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.137,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.1269,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.13,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.1094,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0915,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.1058,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.1011,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0976,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0828,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0861,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0727,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0668,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0653,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0611,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0627,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0566,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0542,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918680877353552e-05,
+      "loss": 5.0321,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0285,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 5.0269,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 5.0268,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 5.0191,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 5.012,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 5.0031,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 5.0107,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9842,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9914,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9879,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.98,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9758,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9802,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.967,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9692,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9623,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9503,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9566,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9542,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9492,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9355,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9415,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9435,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.9307,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.9178,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.9269,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.9257,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.9131,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.908,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8988,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.9032,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8938,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.9078,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8976,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8796,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.887,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.885,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8804,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.884,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8704,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.873,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8786,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8619,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8817958113502494e-05,
+      "loss": 4.8579,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809572165991974e-05,
+      "loss": 4.8597,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8461,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879281664977466e-05,
+      "loss": 4.8452,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878443070226414e-05,
+      "loss": 4.8503,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877604475475362e-05,
+      "loss": 4.8354,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87676588072431e-05,
+      "loss": 4.8393,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875927285973258e-05,
+      "loss": 4.8436,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875088691222206e-05,
+      "loss": 4.8395,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.797691345214844,
+      "eval_runtime": 298.3239,
+      "eval_samples_per_second": 1279.116,
+      "eval_steps_per_second": 39.973,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.8214,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.8208,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872574544849423e-05,
+      "loss": 4.8377,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871735950098371e-05,
+      "loss": 4.8212,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870897355347319e-05,
+      "loss": 4.8322,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870058760596267e-05,
+      "loss": 4.8043,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869220165845215e-05,
+      "loss": 4.8221,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868383208974536e-05,
+      "loss": 4.8045,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7989,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.8059,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.8087,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.8093,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7914,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7907,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.783,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7841,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7822,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7827,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.777,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.7977,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.7717,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.7818,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558092013498754e-05,
+      "loss": 4.7678,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854970606598824e-05,
+      "loss": 4.7768,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854132011847772e-05,
+      "loss": 4.7645,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.7636,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.7584,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851617865474989e-05,
+      "loss": 4.7553,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850779270723937e-05,
+      "loss": 4.7384,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849940675972885e-05,
+      "loss": 4.754,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849102081221833e-05,
+      "loss": 4.7511,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848263486470781e-05,
+      "loss": 4.7543,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847424891719729e-05,
+      "loss": 4.7446,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.747,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.7382,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.7346,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.7193,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.7271,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.7247,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.7233,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.7271,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839880814721007e-05,
+      "loss": 4.7266,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.7149,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838205263099276e-05,
+      "loss": 4.7205,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373666683482236e-05,
+      "loss": 4.7189,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365280735971716e-05,
+      "loss": 4.7052,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356894788461196e-05,
+      "loss": 4.7188,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348508840950676e-05,
+      "loss": 4.706,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.6933,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833175332473337e-05,
+      "loss": 4.7115,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832336737722285e-05,
+      "loss": 4.6952,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831498142971233e-05,
+      "loss": 4.6934,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830659548220181e-05,
+      "loss": 4.682,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829820953469129e-05,
+      "loss": 4.6918,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82898399659845e-05,
+      "loss": 4.6794,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828145401847398e-05,
+      "loss": 4.7121,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827306807096346e-05,
+      "loss": 4.6811,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826468212345294e-05,
+      "loss": 4.6934,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.6897,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.6764,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.6671,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82311547122146e-05,
+      "loss": 4.6787,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222768764704077e-05,
+      "loss": 4.6643,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6631,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6812,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.6768,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189241353465726e-05,
+      "loss": 4.6593,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180855405955206e-05,
+      "loss": 4.654,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172469458444685e-05,
+      "loss": 4.6533,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164083510934165e-05,
+      "loss": 4.6649,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155697563423645e-05,
+      "loss": 4.6636,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147311615913125e-05,
+      "loss": 4.6621,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138942047206334e-05,
+      "loss": 4.6618,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130556099695814e-05,
+      "loss": 4.6632,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122170152185294e-05,
+      "loss": 4.6612,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113784204674774e-05,
+      "loss": 4.6458,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810541463596799e-05,
+      "loss": 4.6493,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809702868845747e-05,
+      "loss": 4.6463,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808864274094695e-05,
+      "loss": 4.6329,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808025679343642e-05,
+      "loss": 4.6478,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80718708459259e-05,
+      "loss": 4.6465,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806348489841538e-05,
+      "loss": 4.6491,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80551153297086e-05,
+      "loss": 4.6288,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804672938219807e-05,
+      "loss": 4.6202,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.6377,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.633,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.638,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.6308,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800481602344921e-05,
+      "loss": 4.6325,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996446454742424e-05,
+      "loss": 4.6266,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79880605072319e-05,
+      "loss": 4.6188,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797967455972138e-05,
+      "loss": 4.6232,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797128861221086e-05,
+      "loss": 4.6226,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.6279,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795451671718982e-05,
+      "loss": 4.6292,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79461307696793e-05,
+      "loss": 4.6255,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937761200972506e-05,
+      "loss": 4.6116,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929375253461986e-05,
+      "loss": 4.6107,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.6093,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.6193,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.6058,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789583146341991e-05,
+      "loss": 4.6046,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5983,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787905956839887e-05,
+      "loss": 4.6122,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787067362088835e-05,
+      "loss": 4.586,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786230405218156e-05,
+      "loss": 4.5988,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785391810467104e-05,
+      "loss": 4.6,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784553215716052e-05,
+      "loss": 4.5937,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783714620965e-05,
+      "loss": 4.5916,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782876026213948e-05,
+      "loss": 4.602,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782037431462896e-05,
+      "loss": 4.5867,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781200474592217e-05,
+      "loss": 4.591,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5942,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795249229704866e-05,
+      "loss": 4.5823,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786863282194346e-05,
+      "loss": 4.5913,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778477334683826e-05,
+      "loss": 4.5965,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770091387173306e-05,
+      "loss": 4.5882,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761705439662786e-05,
+      "loss": 4.5755,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775331949215226e-05,
+      "loss": 4.5883,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744949923445475e-05,
+      "loss": 4.5865,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736563975934955e-05,
+      "loss": 4.5784,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728178028424435e-05,
+      "loss": 4.5726,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771979208091391e-05,
+      "loss": 4.576,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771140613340339e-05,
+      "loss": 4.5879,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703020185892875e-05,
+      "loss": 4.5755,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694634238382355e-05,
+      "loss": 4.5716,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686248290871835e-05,
+      "loss": 4.5619,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7677862343361315e-05,
+      "loss": 4.5685,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7669476395850795e-05,
+      "loss": 4.5636,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7661106827144004e-05,
+      "loss": 4.5792,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652720879633484e-05,
+      "loss": 4.575,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644334932122964e-05,
+      "loss": 4.5551,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763596536341617e-05,
+      "loss": 4.5618,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762757941590565e-05,
+      "loss": 4.5655,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.564,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5707,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760243795217783e-05,
+      "loss": 4.5566,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759405200466731e-05,
+      "loss": 4.5598,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5671,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5533,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.5545,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756050821462523e-05,
+      "loss": 4.5596,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755213864591844e-05,
+      "loss": 4.544,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754375269840792e-05,
+      "loss": 4.5479,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75353667508974e-05,
+      "loss": 4.5547,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752698080338688e-05,
+      "loss": 4.5406,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751859485587636e-05,
+      "loss": 4.5472,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751020890836584e-05,
+      "loss": 4.5504,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750182296085532e-05,
+      "loss": 4.5474,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.510871410369873,
+      "eval_runtime": 296.5965,
+      "eval_samples_per_second": 1286.566,
+      "eval_steps_per_second": 40.206,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74934370133448e-05,
+      "loss": 4.5325,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748505106583428e-05,
+      "loss": 4.5329,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747666511832376e-05,
+      "loss": 4.5536,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746827917081324e-05,
+      "loss": 4.532,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745989322330272e-05,
+      "loss": 4.5484,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74515072757922e-05,
+      "loss": 4.5299,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744312132828168e-05,
+      "loss": 4.5444,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743473538077116e-05,
+      "loss": 4.5256,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426349433260637e-05,
+      "loss": 4.5355,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417963485750117e-05,
+      "loss": 4.5314,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409577538239597e-05,
+      "loss": 4.5356,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401191590729076e-05,
+      "loss": 4.5416,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392838400826015e-05,
+      "loss": 4.5214,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384452453315495e-05,
+      "loss": 4.5249,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376066505804975e-05,
+      "loss": 4.5194,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736768055829446e-05,
+      "loss": 4.5174,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735929461078394e-05,
+      "loss": 4.5201,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735090866327342e-05,
+      "loss": 4.5245,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73425227157629e-05,
+      "loss": 4.5174,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733413676825238e-05,
+      "loss": 4.5415,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732575082074186e-05,
+      "loss": 4.5209,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317364873231334e-05,
+      "loss": 4.5259,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7308978925720814e-05,
+      "loss": 4.5174,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300592978210294e-05,
+      "loss": 4.5269,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292207030699774e-05,
+      "loss": 4.5163,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728383746199298e-05,
+      "loss": 4.5107,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727545151448246e-05,
+      "loss": 4.5148,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726706556697194e-05,
+      "loss": 4.5087,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725869599826516e-05,
+      "loss": 4.4927,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725031005075464e-05,
+      "loss": 4.5096,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724192410324412e-05,
+      "loss": 4.515,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72335381557336e-05,
+      "loss": 4.5109,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722515220822308e-05,
+      "loss": 4.5076,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721676626071256e-05,
+      "loss": 4.5082,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720838031320204e-05,
+      "loss": 4.5017,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719999436569152e-05,
+      "loss": 4.4965,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191608418181e-05,
+      "loss": 4.4929,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718323884947421e-05,
+      "loss": 4.4912,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717485290196369e-05,
+      "loss": 4.4925,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716646695445317e-05,
+      "loss": 4.493,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715808100694265e-05,
+      "loss": 4.4996,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714969505943213e-05,
+      "loss": 4.5,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141325490725343e-05,
+      "loss": 4.4905,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7132939543214823e-05,
+      "loss": 4.4949,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124553595704303e-05,
+      "loss": 4.4985,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711616764819378e-05,
+      "loss": 4.4804,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710779807948699e-05,
+      "loss": 4.4952,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709941213197647e-05,
+      "loss": 4.4883,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709102618446595e-05,
+      "loss": 4.4699,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708264023695543e-05,
+      "loss": 4.4901,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707425428944491e-05,
+      "loss": 4.4821,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706586834193439e-05,
+      "loss": 4.4765,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057482394423865e-05,
+      "loss": 4.4729,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704909644691335e-05,
+      "loss": 4.4769,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704072687820657e-05,
+      "loss": 4.4643,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703235730949978e-05,
+      "loss": 4.4936,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702397136198926e-05,
+      "loss": 4.4736,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701558541447874e-05,
+      "loss": 4.4809,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700719946696822e-05,
+      "loss": 4.4835,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69988135194577e-05,
+      "loss": 4.4683,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699042757194717e-05,
+      "loss": 4.46,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982058003240386e-05,
+      "loss": 4.4751,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973672055729866e-05,
+      "loss": 4.4614,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696528610821934e-05,
+      "loss": 4.4566,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695690016070882e-05,
+      "loss": 4.4767,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948514213198306e-05,
+      "loss": 4.4742,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940128265687786e-05,
+      "loss": 4.4579,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931742318177266e-05,
+      "loss": 4.4532,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923372749470475e-05,
+      "loss": 4.4559,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6914986801959955e-05,
+      "loss": 4.4631,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906600854449435e-05,
+      "loss": 4.4666,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898214906938915e-05,
+      "loss": 4.4658,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6889828959428395e-05,
+      "loss": 4.4672,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881443011917875e-05,
+      "loss": 4.4654,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873057064407355e-05,
+      "loss": 4.469,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864687495700564e-05,
+      "loss": 4.4535,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856301548190044e-05,
+      "loss": 4.4567,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684793197948325e-05,
+      "loss": 4.4591,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683954603197274e-05,
+      "loss": 4.44,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683116008446222e-05,
+      "loss": 4.4567,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68227741369517e-05,
+      "loss": 4.4616,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681438818944118e-05,
+      "loss": 4.4598,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680600224193066e-05,
+      "loss": 4.4453,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679761629442014e-05,
+      "loss": 4.4337,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678923034690962e-05,
+      "loss": 4.4522,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67808443993991e-05,
+      "loss": 4.4412,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677245845188858e-05,
+      "loss": 4.4573,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676408888318179e-05,
+      "loss": 4.4477,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675570293567127e-05,
+      "loss": 4.4503,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674731698816075e-05,
+      "loss": 4.4446,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673893104065023e-05,
+      "loss": 4.4341,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673056147194344e-05,
+      "loss": 4.4425,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722175524432924e-05,
+      "loss": 4.4407,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713789576922404e-05,
+      "loss": 4.4521,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705403629411884e-05,
+      "loss": 4.4543,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669701768190136e-05,
+      "loss": 4.4458,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668863173439084e-05,
+      "loss": 4.4317,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668024578688032e-05,
+      "loss": 4.4367,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66718598393698e-05,
+      "loss": 4.4363,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666347389185928e-05,
+      "loss": 4.4435,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665512070195622e-05,
+      "loss": 4.432,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664675113324943e-05,
+      "loss": 4.4337,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663836518573891e-05,
+      "loss": 4.4241,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662997923822839e-05,
+      "loss": 4.444,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662159329071788e-05,
+      "loss": 4.4153,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.4299,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660482139569683e-05,
+      "loss": 4.4233,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659643544818631e-05,
+      "loss": 4.428,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658804950067579e-05,
+      "loss": 4.423,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657966355316527e-05,
+      "loss": 4.4336,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657129398445848e-05,
+      "loss": 4.4261,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656290803694796e-05,
+      "loss": 4.4212,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655452208943744e-05,
+      "loss": 4.4295,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654613614192692e-05,
+      "loss": 4.4142,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65377501944164e-05,
+      "loss": 4.4331,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652936424690588e-05,
+      "loss": 4.4374,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6520994678199095e-05,
+      "loss": 4.4212,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512608730688575e-05,
+      "loss": 4.4171,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504222783178055e-05,
+      "loss": 4.4229,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495836835667535e-05,
+      "loss": 4.4249,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487450888157015e-05,
+      "loss": 4.417,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479064940646495e-05,
+      "loss": 4.4145,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470678993135975e-05,
+      "loss": 4.4125,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462293045625455e-05,
+      "loss": 4.4302,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6453923476918664e-05,
+      "loss": 4.4198,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445537529408144e-05,
+      "loss": 4.4142,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437151581897624e-05,
+      "loss": 4.4031,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642878201319083e-05,
+      "loss": 4.4142,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642039606568031e-05,
+      "loss": 4.4049,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64120101181698e-05,
+      "loss": 4.4239,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640362417065928e-05,
+      "loss": 4.4196,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639523822314876e-05,
+      "loss": 4.4054,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638685227563824e-05,
+      "loss": 4.402,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637846632812772e-05,
+      "loss": 4.4119,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637008038061719e-05,
+      "loss": 4.4086,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636169443310667e-05,
+      "loss": 4.4213,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635332486439989e-05,
+      "loss": 4.4008,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634493891688936e-05,
+      "loss": 4.4126,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633655296937884e-05,
+      "loss": 4.41,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632816702186832e-05,
+      "loss": 4.4075,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631979745316154e-05,
+      "loss": 4.4074,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631141150565102e-05,
+      "loss": 4.408,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63030255581405e-05,
+      "loss": 4.3974,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629463961062998e-05,
+      "loss": 4.4028,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628627004192319e-05,
+      "loss": 4.4062,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277884094412666e-05,
+      "loss": 4.3987,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269498146902146e-05,
+      "loss": 4.3952,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261112199391626e-05,
+      "loss": 4.4041,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252742630684835e-05,
+      "loss": 4.4015,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.372279167175293,
+      "eval_runtime": 293.5676,
+      "eval_samples_per_second": 1299.84,
+      "eval_steps_per_second": 40.621,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244356683174315e-05,
+      "loss": 4.3951,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6235970735663795e-05,
+      "loss": 4.3849,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227584788153275e-05,
+      "loss": 4.4101,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219215219446484e-05,
+      "loss": 4.3895,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621082927193597e-05,
+      "loss": 4.4094,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620244332442545e-05,
+      "loss": 4.3834,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619405737691493e-05,
+      "loss": 4.4034,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618567142940441e-05,
+      "loss": 4.3876,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617728548189389e-05,
+      "loss": 4.3915,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61689159131871e-05,
+      "loss": 4.394,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616052996567658e-05,
+      "loss": 4.3945,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615214401816606e-05,
+      "loss": 4.4011,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614377444945927e-05,
+      "loss": 4.3847,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613538850194875e-05,
+      "loss": 4.3834,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612700255443823e-05,
+      "loss": 4.3875,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611861660692771e-05,
+      "loss": 4.3737,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611023065941719e-05,
+      "loss": 4.3844,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610184471190667e-05,
+      "loss": 4.3894,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093458764396156e-05,
+      "loss": 4.3802,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085072816885636e-05,
+      "loss": 4.4027,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076686869375116e-05,
+      "loss": 4.3894,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068317300668325e-05,
+      "loss": 4.395,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3826,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051545405647285e-05,
+      "loss": 4.3887,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043159458136764e-05,
+      "loss": 4.383,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6034789889429974e-05,
+      "loss": 4.3789,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026403941919454e-05,
+      "loss": 4.3822,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018017994408933e-05,
+      "loss": 4.3742,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009632046898413e-05,
+      "loss": 4.369,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600126247819162e-05,
+      "loss": 4.3731,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599287653068111e-05,
+      "loss": 4.3829,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598449058317059e-05,
+      "loss": 4.3821,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597610463566007e-05,
+      "loss": 4.3789,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596771868814955e-05,
+      "loss": 4.3818,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595934911944276e-05,
+      "loss": 4.3658,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595096317193224e-05,
+      "loss": 4.3694,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594257722442172e-05,
+      "loss": 4.3693,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59341912769112e-05,
+      "loss": 4.3606,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592580532940067e-05,
+      "loss": 4.3675,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591741938189015e-05,
+      "loss": 4.3642,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590903343437963e-05,
+      "loss": 4.3758,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590066386567285e-05,
+      "loss": 4.3748,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589227791816233e-05,
+      "loss": 4.3656,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588389197065181e-05,
+      "loss": 4.3657,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587550602314129e-05,
+      "loss": 4.3757,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586712007563077e-05,
+      "loss": 4.3581,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858750506923976e-05,
+      "loss": 4.3714,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850364559413456e-05,
+      "loss": 4.3627,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5841978611902936e-05,
+      "loss": 4.347,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833609043196145e-05,
+      "loss": 4.3708,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825223095685625e-05,
+      "loss": 4.3583,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816837148175105e-05,
+      "loss": 4.3583,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808451200664585e-05,
+      "loss": 4.3486,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800065253154065e-05,
+      "loss": 4.3563,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791679305643545e-05,
+      "loss": 4.3447,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783293358133025e-05,
+      "loss": 4.3666,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577490741062251e-05,
+      "loss": 4.356,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576652146311199e-05,
+      "loss": 4.3594,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575813551560147e-05,
+      "loss": 4.365,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574974956809095e-05,
+      "loss": 4.3507,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574136362058043e-05,
+      "loss": 4.3388,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573301043067737e-05,
+      "loss": 4.3605,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.3439,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3409,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570785258814581e-05,
+      "loss": 4.3588,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569946664063529e-05,
+      "loss": 4.3559,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569108069312477e-05,
+      "loss": 4.3434,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568269474561425e-05,
+      "loss": 4.3416,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.334,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665939229396945e-05,
+      "loss": 4.3451,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657553281886425e-05,
+      "loss": 4.3529,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649167334375905e-05,
+      "loss": 4.3534,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640781386865385e-05,
+      "loss": 4.3523,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563239543935486e-05,
+      "loss": 4.3473,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562400949184434e-05,
+      "loss": 4.3611,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615639923137554e-05,
+      "loss": 4.3373,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607253975627034e-05,
+      "loss": 4.3471,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559886802811651e-05,
+      "loss": 4.3443,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559048208060599e-05,
+      "loss": 4.3265,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558209613309547e-05,
+      "loss": 4.3505,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557372656438868e-05,
+      "loss": 4.3485,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556534061687816e-05,
+      "loss": 4.3444,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555695466936764e-05,
+      "loss": 4.3363,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554856872185712e-05,
+      "loss": 4.3276,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55401827743466e-05,
+      "loss": 4.3404,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553179682683608e-05,
+      "loss": 4.3376,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552341087932556e-05,
+      "loss": 4.3449,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551504131061877e-05,
+      "loss": 4.3348,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550667174191198e-05,
+      "loss": 4.345,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549828579440146e-05,
+      "loss": 4.332,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548989984689094e-05,
+      "loss": 4.3325,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548151389938042e-05,
+      "loss": 4.3317,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54731279518699e-05,
+      "loss": 4.329,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546474200435939e-05,
+      "loss": 4.3477,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545635605684887e-05,
+      "loss": 4.3438,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544797010933835e-05,
+      "loss": 4.3448,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543958416182783e-05,
+      "loss": 4.3238,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431214593121036e-05,
+      "loss": 4.3315,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422828645610516e-05,
+      "loss": 4.3325,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414442698099996e-05,
+      "loss": 4.3359,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406056750589476e-05,
+      "loss": 4.3293,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397670803078956e-05,
+      "loss": 4.3263,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389284855568436e-05,
+      "loss": 4.3188,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380898908057916e-05,
+      "loss": 4.3418,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372512960547396e-05,
+      "loss": 4.3121,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364143391840605e-05,
+      "loss": 4.3247,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535577382313382e-05,
+      "loss": 4.3214,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53473878756233e-05,
+      "loss": 4.3244,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533900192811278e-05,
+      "loss": 4.32,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533061598060226e-05,
+      "loss": 4.3331,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532223003309174e-05,
+      "loss": 4.3222,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531384408558122e-05,
+      "loss": 4.3153,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530547451687443e-05,
+      "loss": 4.3321,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.3146,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.3291,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.3373,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527194710563608e-05,
+      "loss": 4.3241,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526356115812556e-05,
+      "loss": 4.3181,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525517521061504e-05,
+      "loss": 4.3174,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524678926310452e-05,
+      "loss": 4.3223,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238403315594e-05,
+      "loss": 4.3191,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523001736808348e-05,
+      "loss": 4.3211,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221647799376694e-05,
+      "loss": 4.3098,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521326185186617e-05,
+      "loss": 4.3309,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520487590435565e-05,
+      "loss": 4.323,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519648995684513e-05,
+      "loss": 4.3183,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518810400933461e-05,
+      "loss": 4.3023,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517971806182409e-05,
+      "loss": 4.3164,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171348493117297e-05,
+      "loss": 4.3096,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5162962545606777e-05,
+      "loss": 4.3218,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154576598096256e-05,
+      "loss": 4.3264,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514619065058574e-05,
+      "loss": 4.3146,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513780470307522e-05,
+      "loss": 4.2995,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512943513436843e-05,
+      "loss": 4.3173,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512104918685791e-05,
+      "loss": 4.3115,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511266323934739e-05,
+      "loss": 4.3268,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510427729183687e-05,
+      "loss": 4.3028,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509589134432635e-05,
+      "loss": 4.316,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508750539681583e-05,
+      "loss": 4.3165,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507911944930531e-05,
+      "loss": 4.3165,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507073350179479e-05,
+      "loss": 4.3113,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062363933088e-05,
+      "loss": 4.3134,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505397798557748e-05,
+      "loss": 4.3031,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50456084168707e-05,
+      "loss": 4.3114,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503722246936018e-05,
+      "loss": 4.3107,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502883652184966e-05,
+      "loss": 4.3082,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502045057433914e-05,
+      "loss": 4.3,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501206462682862e-05,
+      "loss": 4.3134,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50036786793181e-05,
+      "loss": 4.3064,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.286992073059082,
+      "eval_runtime": 294.0694,
+      "eval_samples_per_second": 1297.622,
+      "eval_steps_per_second": 40.552,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995292731807577e-05,
+      "loss": 4.2992,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986906784297057e-05,
+      "loss": 4.2978,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497852083678653e-05,
+      "loss": 4.3161,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497013488927601e-05,
+      "loss": 4.3014,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496174894176549e-05,
+      "loss": 4.3123,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495336299425497e-05,
+      "loss": 4.3002,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494497704674445e-05,
+      "loss": 4.307,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493659109923393e-05,
+      "loss": 4.3003,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492820515172341e-05,
+      "loss": 4.3005,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919819204212896e-05,
+      "loss": 4.3036,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911433256702376e-05,
+      "loss": 4.3054,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903047309191856e-05,
+      "loss": 4.3112,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894677740485065e-05,
+      "loss": 4.2973,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886291792974545e-05,
+      "loss": 4.2953,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877905845464025e-05,
+      "loss": 4.2969,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869519897953505e-05,
+      "loss": 4.2857,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861133950442985e-05,
+      "loss": 4.2961,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852748002932465e-05,
+      "loss": 4.3002,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844362055421945e-05,
+      "loss": 4.2923,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483597610791142e-05,
+      "loss": 4.3143,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48275901604009e-05,
+      "loss": 4.3023,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481920421289038e-05,
+      "loss": 4.3064,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4810818265379865e-05,
+      "loss": 4.295,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802432317869345e-05,
+      "loss": 4.3028,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479407912796629e-05,
+      "loss": 4.293,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478569318045577e-05,
+      "loss": 4.2978,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477730723294524e-05,
+      "loss": 4.2943,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476892128543472e-05,
+      "loss": 4.29,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47605353379242e-05,
+      "loss": 4.286,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475214939041368e-05,
+      "loss": 4.2864,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474376344290316e-05,
+      "loss": 4.2938,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473537749539264e-05,
+      "loss": 4.2978,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472700792668585e-05,
+      "loss": 4.2939,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471862197917533e-05,
+      "loss": 4.2948,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710268789272283e-05,
+      "loss": 4.2815,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701882841761763e-05,
+      "loss": 4.2876,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693496894251243e-05,
+      "loss": 4.2876,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685110946740717e-05,
+      "loss": 4.2753,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676724999230197e-05,
+      "loss": 4.2879,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668339051719676e-05,
+      "loss": 4.278,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4659953104209156e-05,
+      "loss": 4.2925,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651567156698636e-05,
+      "loss": 4.2917,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643181209188116e-05,
+      "loss": 4.2858,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634795261677596e-05,
+      "loss": 4.2806,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626409314167076e-05,
+      "loss": 4.2956,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.2751,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.2908,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460126785043925e-05,
+      "loss": 4.2807,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459288190292873e-05,
+      "loss": 4.2663,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458449595541821e-05,
+      "loss": 4.2911,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457611000790769e-05,
+      "loss": 4.2768,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456772406039717e-05,
+      "loss": 4.2761,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455933811288665e-05,
+      "loss": 4.2718,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455096854417986e-05,
+      "loss": 4.2711,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454258259666934e-05,
+      "loss": 4.2707,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453421302796255e-05,
+      "loss": 4.278,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452582708045203e-05,
+      "loss": 4.2777,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451744113294151e-05,
+      "loss": 4.2779,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450905518543099e-05,
+      "loss": 4.2855,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450066923792047e-05,
+      "loss": 4.2742,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492299669213686e-05,
+      "loss": 4.2577,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483913721703166e-05,
+      "loss": 4.2812,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475527774192646e-05,
+      "loss": 4.2638,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467141826682126e-05,
+      "loss": 4.2599,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458755879171606e-05,
+      "loss": 4.2824,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445036993166108e-05,
+      "loss": 4.278,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444198398415056e-05,
+      "loss": 4.2655,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443359803664004e-05,
+      "loss": 4.2654,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442521208912952e-05,
+      "loss": 4.2538,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416858899226464e-05,
+      "loss": 4.2684,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408472951715944e-05,
+      "loss": 4.2798,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400087004205423e-05,
+      "loss": 4.2741,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391701056694903e-05,
+      "loss": 4.2743,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438331510918438e-05,
+      "loss": 4.271,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437492916167386e-05,
+      "loss": 4.2876,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436654321416334e-05,
+      "loss": 4.2607,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435817364545655e-05,
+      "loss": 4.2735,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434978769794603e-05,
+      "loss": 4.2704,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434140175043551e-05,
+      "loss": 4.2457,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433301580292499e-05,
+      "loss": 4.2768,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432462985541447e-05,
+      "loss": 4.2712,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431624390790395e-05,
+      "loss": 4.2696,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430785796039343e-05,
+      "loss": 4.2637,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429947201288291e-05,
+      "loss": 4.2496,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429110244417613e-05,
+      "loss": 4.2642,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428273287546934e-05,
+      "loss": 4.2627,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427434692795882e-05,
+      "loss": 4.2719,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42659609804483e-05,
+      "loss": 4.26,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425757503293778e-05,
+      "loss": 4.2702,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424918908542726e-05,
+      "loss": 4.2582,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424080313791674e-05,
+      "loss": 4.2623,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423241719040622e-05,
+      "loss": 4.2559,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42240312428957e-05,
+      "loss": 4.2552,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421564529538518e-05,
+      "loss": 4.2727,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207275726678386e-05,
+      "loss": 4.269,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198889779167866e-05,
+      "loss": 4.2746,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190503831657346e-05,
+      "loss": 4.2466,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182117884146826e-05,
+      "loss": 4.2581,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.2604,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416537874673325e-05,
+      "loss": 4.2619,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415699279922273e-05,
+      "loss": 4.2589,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414860685171221e-05,
+      "loss": 4.2495,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414022090420169e-05,
+      "loss": 4.2528,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413183495669117e-05,
+      "loss": 4.2689,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412344900918065e-05,
+      "loss": 4.2393,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411506306167013e-05,
+      "loss": 4.2533,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410667711415961e-05,
+      "loss": 4.2462,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098291166649083e-05,
+      "loss": 4.2584,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.2488,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081552029235515e-05,
+      "loss": 4.2618,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073166081724995e-05,
+      "loss": 4.2535,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064780134214475e-05,
+      "loss": 4.245,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2604,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048024617997164e-05,
+      "loss": 4.2461,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039638670486644e-05,
+      "loss": 4.2609,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.265,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022866775465604e-05,
+      "loss": 4.2522,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014480827955084e-05,
+      "loss": 4.2467,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400609488044456e-05,
+      "loss": 4.2508,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399770893293404e-05,
+      "loss": 4.2495,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398932298542352e-05,
+      "loss": 4.2521,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398096979552047e-05,
+      "loss": 4.2509,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397258384800995e-05,
+      "loss": 4.2417,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396419790049943e-05,
+      "loss": 4.2607,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395581195298891e-05,
+      "loss": 4.2513,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394742600547839e-05,
+      "loss": 4.2522,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393904005796786e-05,
+      "loss": 4.2352,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393065411045734e-05,
+      "loss": 4.2474,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392226816294682e-05,
+      "loss": 4.2409,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391389859424003e-05,
+      "loss": 4.2526,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390551264672951e-05,
+      "loss": 4.2552,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389712669921899e-05,
+      "loss": 4.2499,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388874075170847e-05,
+      "loss": 4.2289,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388035480419795e-05,
+      "loss": 4.2519,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387196885668744e-05,
+      "loss": 4.2451,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386358290917692e-05,
+      "loss": 4.2581,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855213340470127e-05,
+      "loss": 4.2375,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846827392959606e-05,
+      "loss": 4.2533,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838441445449086e-05,
+      "loss": 4.2496,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830071876742296e-05,
+      "loss": 4.2521,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821685929231775e-05,
+      "loss": 4.2429,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813316360524985e-05,
+      "loss": 4.2476,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804930413014465e-05,
+      "loss": 4.2363,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796544465503944e-05,
+      "loss": 4.2453,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788158517993424e-05,
+      "loss": 4.2447,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779772570482904e-05,
+      "loss": 4.242,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377140300177612e-05,
+      "loss": 4.2358,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37630170542656e-05,
+      "loss": 4.2452,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375463110675508e-05,
+      "loss": 4.245,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.22967004776001,
+      "eval_runtime": 293.6171,
+      "eval_samples_per_second": 1299.621,
+      "eval_steps_per_second": 40.614,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374624515924456e-05,
+      "loss": 4.2396,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373785921173404e-05,
+      "loss": 4.2307,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372947326422352e-05,
+      "loss": 4.2493,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721087316713e-05,
+      "loss": 4.2374,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371270136920248e-05,
+      "loss": 4.2455,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370431542169196e-05,
+      "loss": 4.2383,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369592947418144e-05,
+      "loss": 4.2391,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368754352667092e-05,
+      "loss": 4.2398,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367917395796413e-05,
+      "loss": 4.2347,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367078801045361e-05,
+      "loss": 4.2386,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366240206294309e-05,
+      "loss": 4.2453,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654016115432576e-05,
+      "loss": 4.2421,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645646546725785e-05,
+      "loss": 4.2338,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637260599215265e-05,
+      "loss": 4.2308,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3628874651704745e-05,
+      "loss": 4.2363,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620488704194225e-05,
+      "loss": 4.2194,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36121027566837e-05,
+      "loss": 4.233,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360371680917318e-05,
+      "loss": 4.2339,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359533086166266e-05,
+      "loss": 4.2336,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358694491415214e-05,
+      "loss": 4.2512,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357857534544535e-05,
+      "loss": 4.2407,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357018939793483e-05,
+      "loss": 4.2435,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356181982922804e-05,
+      "loss": 4.2334,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355343388171752e-05,
+      "loss": 4.2383,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545047934207e-05,
+      "loss": 4.2318,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353666198669648e-05,
+      "loss": 4.2344,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352827603918596e-05,
+      "loss": 4.2303,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351989009167544e-05,
+      "loss": 4.229,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351150414416492e-05,
+      "loss": 4.2257,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350313457545813e-05,
+      "loss": 4.2262,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349474862794761e-05,
+      "loss": 4.2302,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348636268043709e-05,
+      "loss": 4.2347,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347797673292657e-05,
+      "loss": 4.2305,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.2365,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.2199,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345281889039501e-05,
+      "loss": 4.2277,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344443294288449e-05,
+      "loss": 4.2264,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343606337417771e-05,
+      "loss": 4.2155,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342767742666719e-05,
+      "loss": 4.2296,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341929147915667e-05,
+      "loss": 4.2134,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341090553164615e-05,
+      "loss": 4.2316,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402535962939356e-05,
+      "loss": 4.2307,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394166394232565e-05,
+      "loss": 4.2294,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385780446722045e-05,
+      "loss": 4.2209,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377394499211525e-05,
+      "loss": 4.2365,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369008551701005e-05,
+      "loss": 4.2134,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360622604190485e-05,
+      "loss": 4.2334,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352236656679965e-05,
+      "loss": 4.2207,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343850709169445e-05,
+      "loss": 4.201,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333548114046266e-05,
+      "loss": 4.2359,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332709519295214e-05,
+      "loss": 4.2147,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331870924544162e-05,
+      "loss": 4.222,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.2097,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.2105,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.2162,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285165455399534e-05,
+      "loss": 4.2169,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276779507889013e-05,
+      "loss": 4.2225,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.218,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.2313,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.2138,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324325209665067e-05,
+      "loss": 4.2031,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323486614914015e-05,
+      "loss": 4.2225,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322649658043336e-05,
+      "loss": 4.203,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321811063292284e-05,
+      "loss": 4.2038,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320972468541232e-05,
+      "loss": 4.2251,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32013387379018e-05,
+      "loss": 4.2219,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319296916919501e-05,
+      "loss": 4.2069,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.2108,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.1985,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.2096,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.2209,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.2182,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314266986293562e-05,
+      "loss": 4.2178,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31342839154251e-05,
+      "loss": 4.2147,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312589796791458e-05,
+      "loss": 4.2313,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311751202040406e-05,
+      "loss": 4.2035,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.2166,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310075650418675e-05,
+      "loss": 4.2123,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309237055667623e-05,
+      "loss": 4.1927,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308398460916571e-05,
+      "loss": 4.2188,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307561504045892e-05,
+      "loss": 4.2164,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30672290929484e-05,
+      "loss": 4.2149,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.2104,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305045719792736e-05,
+      "loss": 4.1985,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.2033,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.208,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30252993553958e-05,
+      "loss": 4.2158,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016929786689016e-05,
+      "loss": 4.2038,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.2148,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000190649275435e-05,
+      "loss": 4.2027,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991804701764915e-05,
+      "loss": 4.208,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983418754254395e-05,
+      "loss": 4.2004,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975032806743874e-05,
+      "loss": 4.2022,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966646859233354e-05,
+      "loss": 4.2172,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.216,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.2195,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941489016701794e-05,
+      "loss": 4.1925,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933103069191274e-05,
+      "loss": 4.206,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924717121680754e-05,
+      "loss": 4.202,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291633117417024e-05,
+      "loss": 4.211,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290796160546345e-05,
+      "loss": 4.2022,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.2016,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.1955,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288282014173562e-05,
+      "loss": 4.2121,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28744341942251e-05,
+      "loss": 4.1874,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286604824671458e-05,
+      "loss": 4.2033,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285766229920406e-05,
+      "loss": 4.1884,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284927635169354e-05,
+      "loss": 4.2066,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284089040418301e-05,
+      "loss": 4.1986,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283250445667249e-05,
+      "loss": 4.2023,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282411850916198e-05,
+      "loss": 4.1989,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281573256165146e-05,
+      "loss": 4.1989,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280734661414094e-05,
+      "loss": 4.203,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279896066663042e-05,
+      "loss": 4.1965,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27905747191199e-05,
+      "loss": 4.203,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278218877160938e-05,
+      "loss": 4.2171,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277381920290259e-05,
+      "loss": 4.1994,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276543325539207e-05,
+      "loss": 4.194,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275704730788155e-05,
+      "loss": 4.1987,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274866136037103e-05,
+      "loss": 4.2008,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274027541286051e-05,
+      "loss": 4.1915,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273188946534999e-05,
+      "loss": 4.2013,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272350351783947e-05,
+      "loss": 4.188,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715133949132676e-05,
+      "loss": 4.2069,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270674800162216e-05,
+      "loss": 4.2038,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269836205411164e-05,
+      "loss": 4.1956,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268997610660112e-05,
+      "loss": 4.188,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26815901590906e-05,
+      "loss": 4.1934,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267322059038381e-05,
+      "loss": 4.1938,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266483464287329e-05,
+      "loss": 4.2032,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265644869536277e-05,
+      "loss": 4.1995,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264806274785225e-05,
+      "loss": 4.2015,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639676800341725e-05,
+      "loss": 4.1777,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631290852831205e-05,
+      "loss": 4.2008,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2622904905320685e-05,
+      "loss": 4.1952,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614518957810165e-05,
+      "loss": 4.2064,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260614938910338e-05,
+      "loss": 4.1867,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259776344159286e-05,
+      "loss": 4.2056,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258937749408234e-05,
+      "loss": 4.1968,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258099154657182e-05,
+      "loss": 4.2028,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257262197786503e-05,
+      "loss": 4.1883,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256423603035451e-05,
+      "loss": 4.2042,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255585008284399e-05,
+      "loss": 4.18,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254746413533347e-05,
+      "loss": 4.1965,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253909456662668e-05,
+      "loss": 4.1954,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253070861911616e-05,
+      "loss": 4.1869,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252232267160564e-05,
+      "loss": 4.193,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251393672409512e-05,
+      "loss": 4.1888,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250558353419207e-05,
+      "loss": 4.1964,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.18727445602417,
+      "eval_runtime": 300.1993,
+      "eval_samples_per_second": 1271.126,
+      "eval_steps_per_second": 39.724,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249719758668155e-05,
+      "loss": 4.1922,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488811639171024e-05,
+      "loss": 4.1789,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480425691660504e-05,
+      "loss": 4.1966,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472039744149984e-05,
+      "loss": 4.1871,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463653796639464e-05,
+      "loss": 4.1965,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455267849128943e-05,
+      "loss": 4.1874,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2446881901618423e-05,
+      "loss": 4.1902,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24384959541079e-05,
+      "loss": 4.1919,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243011000659738e-05,
+      "loss": 4.1868,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242172405908686e-05,
+      "loss": 4.1895,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241333811157634e-05,
+      "loss": 4.1947,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240495216406582e-05,
+      "loss": 4.1938,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239658259535903e-05,
+      "loss": 4.1858,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238819664784852e-05,
+      "loss": 4.1829,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379810700338e-05,
+      "loss": 4.1876,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237142475282748e-05,
+      "loss": 4.1692,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236303880531696e-05,
+      "loss": 4.1856,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235465285780644e-05,
+      "loss": 4.1851,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234626691029591e-05,
+      "loss": 4.1808,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233788096278539e-05,
+      "loss": 4.2039,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232949501527487e-05,
+      "loss": 4.1956,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232112544656809e-05,
+      "loss": 4.1971,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231273949905756e-05,
+      "loss": 4.1835,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230435355154704e-05,
+      "loss": 4.1899,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229596760403652e-05,
+      "loss": 4.1855,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287581656526e-05,
+      "loss": 4.1858,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227919570901549e-05,
+      "loss": 4.1832,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22708261403087e-05,
+      "loss": 4.185,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226244019279818e-05,
+      "loss": 4.1757,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254054245287657e-05,
+      "loss": 4.1783,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245668297777137e-05,
+      "loss": 4.1809,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237282350266617e-05,
+      "loss": 4.1897,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228896402756096e-05,
+      "loss": 4.1833,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220510455245576e-05,
+      "loss": 4.1897,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212124507735056e-05,
+      "loss": 4.173,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203754939028265e-05,
+      "loss": 4.1779,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195368991517745e-05,
+      "loss": 4.1812,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2186983044007225e-05,
+      "loss": 4.1701,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178597096496705e-05,
+      "loss": 4.1782,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170211148986185e-05,
+      "loss": 4.1709,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216182520147567e-05,
+      "loss": 4.181,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215343925396515e-05,
+      "loss": 4.183,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145053306454625e-05,
+      "loss": 4.1794,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136667358944105e-05,
+      "loss": 4.1762,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128281411433585e-05,
+      "loss": 4.186,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21199118427268e-05,
+      "loss": 4.1714,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111525895216274e-05,
+      "loss": 4.1838,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103139947705754e-05,
+      "loss": 4.1758,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094754000195234e-05,
+      "loss": 4.1529,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086368052684714e-05,
+      "loss": 4.1975,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207799848397792e-05,
+      "loss": 4.1676,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206961253646741e-05,
+      "loss": 4.1788,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206122658895689e-05,
+      "loss": 4.1606,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205284064144637e-05,
+      "loss": 4.1689,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204447107273958e-05,
+      "loss": 4.1644,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036101504032795e-05,
+      "loss": 4.1734,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027715556522275e-05,
+      "loss": 4.1749,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201932960901175e-05,
+      "loss": 4.1757,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201094366150123e-05,
+      "loss": 4.1838,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200255771399071e-05,
+      "loss": 4.1696,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199417176648019e-05,
+      "loss": 4.1563,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198578581896967e-05,
+      "loss": 4.1826,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197739987145915e-05,
+      "loss": 4.1574,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969030302752363e-05,
+      "loss": 4.1559,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1960644355241843e-05,
+      "loss": 4.1809,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952258407731323e-05,
+      "loss": 4.1781,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19438724602208e-05,
+      "loss": 4.1603,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193550289151401e-05,
+      "loss": 4.1695,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192711694400349e-05,
+      "loss": 4.1532,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191873099649297e-05,
+      "loss": 4.1656,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191034504898245e-05,
+      "loss": 4.174,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190195910147193e-05,
+      "loss": 4.1719,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189358953276514e-05,
+      "loss": 4.1759,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188520358525462e-05,
+      "loss": 4.1733,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18768176377441e-05,
+      "loss": 4.1835,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186843169023358e-05,
+      "loss": 4.1579,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186004574272306e-05,
+      "loss": 4.174,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185165979521254e-05,
+      "loss": 4.1647,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184327384770203e-05,
+      "loss": 4.1549,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183488790019151e-05,
+      "loss": 4.1692,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182651833148472e-05,
+      "loss": 4.1756,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18181323839742e-05,
+      "loss": 4.1717,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180974643646368e-05,
+      "loss": 4.1685,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180136048895316e-05,
+      "loss": 4.1544,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1792990920246366e-05,
+      "loss": 4.156,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784604972735846e-05,
+      "loss": 4.1689,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776219025225326e-05,
+      "loss": 4.1715,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767833077714806e-05,
+      "loss": 4.1603,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759463509008015e-05,
+      "loss": 4.1724,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751077561497495e-05,
+      "loss": 4.1518,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174269161398698e-05,
+      "loss": 4.168,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173430566647646e-05,
+      "loss": 4.1583,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725919718965935e-05,
+      "loss": 4.1611,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1717533771455415e-05,
+      "loss": 4.1746,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709147823944895e-05,
+      "loss": 4.174,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1700761876434375e-05,
+      "loss": 4.1764,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692375928923855e-05,
+      "loss": 4.1498,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684006360217064e-05,
+      "loss": 4.1644,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1675620412706544e-05,
+      "loss": 4.1563,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166725084399975e-05,
+      "loss": 4.1692,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165886489648923e-05,
+      "loss": 4.1596,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165049532778245e-05,
+      "loss": 4.1628,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642109380271935e-05,
+      "loss": 4.1547,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163372343276141e-05,
+      "loss": 4.1666,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162533748525089e-05,
+      "loss": 4.1489,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161695153774037e-05,
+      "loss": 4.1622,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160856559022985e-05,
+      "loss": 4.1458,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160017964271933e-05,
+      "loss": 4.1663,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159179369520881e-05,
+      "loss": 4.1571,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158340774769829e-05,
+      "loss": 4.1593,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157502180018777e-05,
+      "loss": 4.1567,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156663585267725e-05,
+      "loss": 4.1606,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155824990516673e-05,
+      "loss": 4.1596,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154988033645994e-05,
+      "loss": 4.1542,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154149438894942e-05,
+      "loss": 4.1652,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533108441438904e-05,
+      "loss": 4.1715,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152475525153584e-05,
+      "loss": 4.1645,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151636930402532e-05,
+      "loss": 4.1545,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15079833565148e-05,
+      "loss": 4.1542,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149959740900428e-05,
+      "loss": 4.163,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149121146149376e-05,
+      "loss": 4.1501,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148284189278697e-05,
+      "loss": 4.1623,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147445594527645e-05,
+      "loss": 4.1466,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146606999776593e-05,
+      "loss": 4.1667,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145768405025541e-05,
+      "loss": 4.1607,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144929810274489e-05,
+      "loss": 4.1578,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144091215523437e-05,
+      "loss": 4.1494,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143252620772386e-05,
+      "loss": 4.1535,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142414026021334e-05,
+      "loss": 4.1529,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141575431270282e-05,
+      "loss": 4.166,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14073683651923e-05,
+      "loss": 4.1588,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139898241768177e-05,
+      "loss": 4.1588,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139059647017125e-05,
+      "loss": 4.1411,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138221052266073e-05,
+      "loss": 4.1615,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373840953953946e-05,
+      "loss": 4.1532,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136545500644342e-05,
+      "loss": 4.1671,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357085437736635e-05,
+      "loss": 4.1477,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348699490226115e-05,
+      "loss": 4.163,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340313542715595e-05,
+      "loss": 4.1544,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1331927595205075e-05,
+      "loss": 4.1692,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323541647694555e-05,
+      "loss": 4.1487,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131517207898777e-05,
+      "loss": 4.1574,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306786131477244e-05,
+      "loss": 4.1442,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298400183966724e-05,
+      "loss": 4.1559,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290014236456204e-05,
+      "loss": 4.1559,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281628288945684e-05,
+      "loss": 4.1467,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273242341435164e-05,
+      "loss": 4.1566,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1264856393924644e-05,
+      "loss": 4.1444,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1256470446414124e-05,
+      "loss": 4.1603,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.156093120574951,
+      "eval_runtime": 305.6676,
+      "eval_samples_per_second": 1248.386,
+      "eval_steps_per_second": 39.013,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124810087770733e-05,
+      "loss": 4.1532,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123971493019681e-05,
+      "loss": 4.14,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123132898268629e-05,
+      "loss": 4.1582,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122294303517577e-05,
+      "loss": 4.1465,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121455708766526e-05,
+      "loss": 4.1594,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120618751895847e-05,
+      "loss": 4.1512,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119780157144795e-05,
+      "loss": 4.1527,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118941562393743e-05,
+      "loss": 4.1543,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118102967642691e-05,
+      "loss": 4.1455,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117266010772012e-05,
+      "loss": 4.1516,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11642741602096e-05,
+      "loss": 4.1553,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.1539,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114751864399229e-05,
+      "loss": 4.1524,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113913269648177e-05,
+      "loss": 4.1408,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1130746748971247e-05,
+      "loss": 4.1446,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122360801460727e-05,
+      "loss": 4.1349,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.1453,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110558890643969e-05,
+      "loss": 4.1506,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.1403,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108881701141865e-05,
+      "loss": 4.1629,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108044744271186e-05,
+      "loss": 4.1583,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107206149520134e-05,
+      "loss": 4.159,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106369192649455e-05,
+      "loss": 4.1473,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105530597898403e-05,
+      "loss": 4.1468,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104692003147351e-05,
+      "loss": 4.1529,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103853408396299e-05,
+      "loss": 4.1436,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103014813645247e-05,
+      "loss": 4.1449,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102177856774568e-05,
+      "loss": 4.152,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101339262023517e-05,
+      "loss": 4.1338,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100500667272465e-05,
+      "loss": 4.1403,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099662072521413e-05,
+      "loss": 4.1437,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098823477770361e-05,
+      "loss": 4.148,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097984883019308e-05,
+      "loss": 4.1491,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097146288268256e-05,
+      "loss": 4.1527,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096307693517204e-05,
+      "loss": 4.1358,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095470736646525e-05,
+      "loss": 4.1381,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094632141895473e-05,
+      "loss": 4.147,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093793547144421e-05,
+      "loss": 4.1336,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092954952393369e-05,
+      "loss": 4.1413,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092116357642317e-05,
+      "loss": 4.1308,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091277762891265e-05,
+      "loss": 4.1464,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904391681402136e-05,
+      "loss": 4.1416,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896005733891615e-05,
+      "loss": 4.1444,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0887636165184825e-05,
+      "loss": 4.143,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879250217674305e-05,
+      "loss": 4.1454,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0870864270163784e-05,
+      "loss": 4.1369,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0862478322653264e-05,
+      "loss": 4.1445,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854108753946474e-05,
+      "loss": 4.139,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084573918523968e-05,
+      "loss": 4.1187,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083735323772916e-05,
+      "loss": 4.1586,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082896729021864e-05,
+      "loss": 4.1295,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082058134270812e-05,
+      "loss": 4.1421,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08121953951976e-05,
+      "loss": 4.1292,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080380944768708e-05,
+      "loss": 4.1292,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079542350017657e-05,
+      "loss": 4.1254,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078703755266605e-05,
+      "loss": 4.1351,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077865160515553e-05,
+      "loss": 4.1409,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077028203644874e-05,
+      "loss": 4.1344,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076189608893822e-05,
+      "loss": 4.1489,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07535101414277e-05,
+      "loss": 4.1401,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074512419391718e-05,
+      "loss": 4.1161,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073673824640666e-05,
+      "loss": 4.1421,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072836867769987e-05,
+      "loss": 4.1244,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071998273018935e-05,
+      "loss": 4.1175,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071159678267883e-05,
+      "loss": 4.1484,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703227213972036e-05,
+      "loss": 4.1394,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069484126646152e-05,
+      "loss": 4.1252,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686455318951e-05,
+      "loss": 4.1343,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067806937144048e-05,
+      "loss": 4.1195,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066969980273369e-05,
+      "loss": 4.1298,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066131385522317e-05,
+      "loss": 4.1365,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065292790771265e-05,
+      "loss": 4.1376,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064454196020213e-05,
+      "loss": 4.1407,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063615601269161e-05,
+      "loss": 4.1345,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627770065181085e-05,
+      "loss": 4.1493,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619384117670565e-05,
+      "loss": 4.1182,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0610998170160045e-05,
+      "loss": 4.1414,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602612222649525e-05,
+      "loss": 4.1265,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594226275139005e-05,
+      "loss": 4.1225,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058584032762849e-05,
+      "loss": 4.1349,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057745438011797e-05,
+      "loss": 4.1397,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056908481141118e-05,
+      "loss": 4.1351,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056069886390066e-05,
+      "loss": 4.1377,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055231291639014e-05,
+      "loss": 4.1151,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054392696887962e-05,
+      "loss": 4.1207,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053555740017283e-05,
+      "loss": 4.1319,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052718783146604e-05,
+      "loss": 4.1378,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051880188395552e-05,
+      "loss": 4.1285,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0510415936445e-05,
+      "loss": 4.1365,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050202998893448e-05,
+      "loss": 4.1183,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049364404142396e-05,
+      "loss": 4.134,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485258093913445e-05,
+      "loss": 4.124,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0476872146402925e-05,
+      "loss": 4.1238,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468502577696134e-05,
+      "loss": 4.1379,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460116630185614e-05,
+      "loss": 4.1377,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0451730682675094e-05,
+      "loss": 4.1418,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443344735164574e-05,
+      "loss": 4.1207,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043497516645778e-05,
+      "loss": 4.1265,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042658921894726e-05,
+      "loss": 4.1244,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041820327143674e-05,
+      "loss": 4.1316,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040981732392622e-05,
+      "loss": 4.1259,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040144775521943e-05,
+      "loss": 4.1303,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039306180770891e-05,
+      "loss": 4.1212,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03846758601984e-05,
+      "loss": 4.1272,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037628991268788e-05,
+      "loss": 4.1221,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036790396517736e-05,
+      "loss": 4.1243,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035951801766684e-05,
+      "loss": 4.1097,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035113207015632e-05,
+      "loss": 4.1333,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03427461226458e-05,
+      "loss": 4.1247,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033437655393901e-05,
+      "loss": 4.1271,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032599060642849e-05,
+      "loss": 4.1215,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031760465891797e-05,
+      "loss": 4.1262,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030921871140745e-05,
+      "loss": 4.1279,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030083276389692e-05,
+      "loss": 4.1206,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0292479573993866e-05,
+      "loss": 4.1312,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028409362648335e-05,
+      "loss": 4.1364,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027570767897283e-05,
+      "loss": 4.1253,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026732173146231e-05,
+      "loss": 4.1289,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025893578395179e-05,
+      "loss": 4.118,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025054983644127e-05,
+      "loss": 4.1242,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242163888930745e-05,
+      "loss": 4.1248,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023379432022396e-05,
+      "loss": 4.1276,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022540837271344e-05,
+      "loss": 4.1109,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021703880400665e-05,
+      "loss": 4.1311,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020865285649613e-05,
+      "loss": 4.1275,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020026690898561e-05,
+      "loss": 4.126,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019188096147509e-05,
+      "loss": 4.1159,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018349501396457e-05,
+      "loss": 4.1199,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017510906645405e-05,
+      "loss": 4.1187,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016672311894353e-05,
+      "loss": 4.1296,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015833717143301e-05,
+      "loss": 4.1246,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014995122392249e-05,
+      "loss": 4.1276,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01415816552157e-05,
+      "loss": 4.1063,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013319570770518e-05,
+      "loss": 4.1277,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012480976019466e-05,
+      "loss": 4.1224,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011642381268414e-05,
+      "loss": 4.133,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010805424397735e-05,
+      "loss": 4.1171,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009966829646683e-05,
+      "loss": 4.1258,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009128234895631e-05,
+      "loss": 4.1207,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008289640144579e-05,
+      "loss": 4.1377,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007451045393527e-05,
+      "loss": 4.1167,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0066124506424755e-05,
+      "loss": 4.1251,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0057738558914235e-05,
+      "loss": 4.1177,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0049352611403714e-05,
+      "loss": 4.1236,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0040983042696924e-05,
+      "loss": 4.1187,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0032597095186404e-05,
+      "loss": 4.1154,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024211147675883e-05,
+      "loss": 4.1235,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0015825200165363e-05,
+      "loss": 4.1112,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0007439252654843e-05,
+      "loss": 4.1312,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.131312370300293,
+      "eval_runtime": 307.6483,
+      "eval_samples_per_second": 1240.348,
+      "eval_steps_per_second": 38.762,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999905330514432e-05,
+      "loss": 4.1158,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99906673576338e-05,
+      "loss": 4.1105,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998228141012328e-05,
+      "loss": 4.1227,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9973895462612756e-05,
+      "loss": 4.1174,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9965509515102236e-05,
+      "loss": 4.1285,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995712356759172e-05,
+      "loss": 4.1187,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99487376200812e-05,
+      "loss": 4.1211,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994035167257068e-05,
+      "loss": 4.123,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993196572506016e-05,
+      "loss": 4.1125,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992357977754964e-05,
+      "loss": 4.121,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991519383003912e-05,
+      "loss": 4.1246,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99068078825286e-05,
+      "loss": 4.1228,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989843831382181e-05,
+      "loss": 4.1243,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989005236631129e-05,
+      "loss": 4.1065,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988166641880077e-05,
+      "loss": 4.1165,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987328047129025e-05,
+      "loss": 4.1022,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986489452377973e-05,
+      "loss": 4.1173,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985650857626921e-05,
+      "loss": 4.1173,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984812262875869e-05,
+      "loss": 4.1097,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983973668124817e-05,
+      "loss": 4.1304,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983135073373765e-05,
+      "loss": 4.1268,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982296478622713e-05,
+      "loss": 4.1256,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981457883871661e-05,
+      "loss": 4.119,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980619289120609e-05,
+      "loss": 4.1179,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979780694369557e-05,
+      "loss": 4.1238,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978943737498878e-05,
+      "loss": 4.1105,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978105142747826e-05,
+      "loss": 4.1123,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977266547996774e-05,
+      "loss": 4.1198,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976429591126095e-05,
+      "loss": 4.1054,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975590996375043e-05,
+      "loss": 4.1076,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974752401623991e-05,
+      "loss": 4.1151,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973913806872939e-05,
+      "loss": 4.1183,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973075212121887e-05,
+      "loss": 4.112,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9722366173708356e-05,
+      "loss": 4.1242,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9713980226197836e-05,
+      "loss": 4.1058,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705610657491045e-05,
+      "loss": 4.1075,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697224709980525e-05,
+      "loss": 4.1138,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9688838762470005e-05,
+      "loss": 4.098,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680452814959485e-05,
+      "loss": 4.1128,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672066867448965e-05,
+      "loss": 4.1032,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663680919938445e-05,
+      "loss": 4.1112,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655294972427925e-05,
+      "loss": 4.1124,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9646909024917405e-05,
+      "loss": 4.1128,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638539456210614e-05,
+      "loss": 4.1146,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9630153508700094e-05,
+      "loss": 4.1137,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621767561189574e-05,
+      "loss": 4.1083,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613381613679054e-05,
+      "loss": 4.1121,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960499566616854e-05,
+      "loss": 4.111,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959662609746175e-05,
+      "loss": 4.0875,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958824014995123e-05,
+      "loss": 4.1226,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957985420244071e-05,
+      "loss": 4.0996,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957146825493018e-05,
+      "loss": 4.1142,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956308230741966e-05,
+      "loss": 4.0997,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955471273871288e-05,
+      "loss": 4.0982,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954632679120236e-05,
+      "loss": 4.0997,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953794084369183e-05,
+      "loss": 4.0997,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952955489618131e-05,
+      "loss": 4.1147,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952116894867079e-05,
+      "loss": 4.0998,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951279937996401e-05,
+      "loss": 4.1213,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950441343245349e-05,
+      "loss": 4.1092,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949602748494297e-05,
+      "loss": 4.087,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948765791623618e-05,
+      "loss": 4.1079,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479271968725656e-05,
+      "loss": 4.1013,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9470886021215136e-05,
+      "loss": 4.0877,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462500073704616e-05,
+      "loss": 4.1148,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454114126194096e-05,
+      "loss": 4.1079,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445728178683576e-05,
+      "loss": 4.096,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437342231173056e-05,
+      "loss": 4.1058,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9428956283662536e-05,
+      "loss": 4.0862,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420586714955745e-05,
+      "loss": 4.1014,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941220076744523e-05,
+      "loss": 4.1056,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940381481993471e-05,
+      "loss": 4.1042,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939542887242419e-05,
+      "loss": 4.1143,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938704292491367e-05,
+      "loss": 4.1006,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937867335620688e-05,
+      "loss": 4.1184,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937028740869636e-05,
+      "loss": 4.0967,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936191783998957e-05,
+      "loss": 4.1091,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935353189247905e-05,
+      "loss": 4.0965,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934514594496853e-05,
+      "loss": 4.0945,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933675999745801e-05,
+      "loss": 4.1048,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932839042875122e-05,
+      "loss": 4.1132,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93200044812407e-05,
+      "loss": 4.1037,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311618533730186e-05,
+      "loss": 4.1093,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303232586219666e-05,
+      "loss": 4.087,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9294846638709146e-05,
+      "loss": 4.0908,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286460691198626e-05,
+      "loss": 4.1062,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278074743688105e-05,
+      "loss": 4.1047,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9269688796177585e-05,
+      "loss": 4.0974,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261302848667065e-05,
+      "loss": 4.1069,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9252916901156545e-05,
+      "loss": 4.0919,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924453095364602e-05,
+      "loss": 4.1022,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236161384939234e-05,
+      "loss": 4.0959,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9227775437428714e-05,
+      "loss": 4.0968,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219389489918194e-05,
+      "loss": 4.1055,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921100354240767e-05,
+      "loss": 4.1089,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920261759489715e-05,
+      "loss": 4.1139,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919424802619036e-05,
+      "loss": 4.092,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918587845748358e-05,
+      "loss": 4.0997,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917749250997306e-05,
+      "loss": 4.0955,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916910656246254e-05,
+      "loss": 4.1017,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916072061495202e-05,
+      "loss": 4.0991,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915233466744149e-05,
+      "loss": 4.0998,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914394871993097e-05,
+      "loss": 4.0968,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913557915122419e-05,
+      "loss": 4.0956,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912719320371367e-05,
+      "loss": 4.0929,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911882363500688e-05,
+      "loss": 4.0957,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911043768749636e-05,
+      "loss": 4.0829,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910205173998584e-05,
+      "loss": 4.1021,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909366579247532e-05,
+      "loss": 4.0981,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90852798449648e-05,
+      "loss": 4.093,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907689389745428e-05,
+      "loss": 4.0972,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906850794994376e-05,
+      "loss": 4.0943,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906012200243324e-05,
+      "loss": 4.0987,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905173605492272e-05,
+      "loss": 4.092,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90433501074122e-05,
+      "loss": 4.1061,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903496415990168e-05,
+      "loss": 4.107,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026578212391157e-05,
+      "loss": 4.0998,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018208643684366e-05,
+      "loss": 4.0981,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9009822696173846e-05,
+      "loss": 4.0925,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9001436748663326e-05,
+      "loss": 4.0923,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899306717995654e-05,
+      "loss": 4.0984,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898468123244602e-05,
+      "loss": 4.0993,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89762952849355e-05,
+      "loss": 4.0814,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896790933742498e-05,
+      "loss": 4.1063,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895952338991446e-05,
+      "loss": 4.0994,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895113744240394e-05,
+      "loss": 4.0989,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894275149489342e-05,
+      "loss": 4.0891,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89343655473829e-05,
+      "loss": 4.0909,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892597959987238e-05,
+      "loss": 4.0914,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917593652361854e-05,
+      "loss": 4.1013,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909207704851334e-05,
+      "loss": 4.0975,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8900821757340814e-05,
+      "loss": 4.1019,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889245218863403e-05,
+      "loss": 4.0773,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888406624112351e-05,
+      "loss": 4.101,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887568029361299e-05,
+      "loss": 4.0958,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867310724906206e-05,
+      "loss": 4.1087,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885892477739568e-05,
+      "loss": 4.0908,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885053882988516e-05,
+      "loss": 4.0943,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884215288237464e-05,
+      "loss": 4.0977,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883376693486412e-05,
+      "loss": 4.1058,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88253809873536e-05,
+      "loss": 4.0886,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881701141864681e-05,
+      "loss": 4.0988,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880862547113629e-05,
+      "loss": 4.0927,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880023952362577e-05,
+      "loss": 4.1011,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879185357611525e-05,
+      "loss": 4.089,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878346762860473e-05,
+      "loss": 4.086,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877508168109421e-05,
+      "loss": 4.0944,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8766695733583694e-05,
+      "loss": 4.0869,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758309786073174e-05,
+      "loss": 4.102,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.111427307128906,
+      "eval_runtime": 303.002,
+      "eval_samples_per_second": 1259.368,
+      "eval_steps_per_second": 39.356,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.8223721053402656e+17,
+  "trial_name": null,
+  "trial_params": null
+}