diff --git "a/checkpoint-1450080/trainer_state.json" "b/checkpoint-1450080/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1450080/trainer_state.json"
@@ -0,0 +1,17169 @@
+{
+  "best_metric": 4.051382541656494,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/lstm/0/checkpoints/checkpoint-1450080",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1450080,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8195,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5629,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0621,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9908,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9369,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8371,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7009,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.5988,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5058,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4276,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3668,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.2887,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2278,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1552,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1049,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.04,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0126,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 5.9574,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9143,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.8808,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8575,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8109,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.7867,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7509,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7184,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7003,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6629,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.6507,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.625,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.5976,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5835,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.561,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5482,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.529,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5105,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5025,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.495,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4556,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 5.4478,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4339,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4193,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4053,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4073,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3857,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.371,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3638,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3546,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3485,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3232,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3127,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3212,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.31,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.2974,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2747,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2834,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.26,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2779,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.243,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2466,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2271,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2167,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2181,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948020228477761e-05,
+      "loss": 5.2027,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9471816337267086e-05,
+      "loss": 5.1976,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1888,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.1986,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1712,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9438288926028735e-05,
+      "loss": 5.1614,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9429902978518215e-05,
+      "loss": 5.1454,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1456,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1594,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404761514790384e-05,
+      "loss": 5.1494,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.939637556727987e-05,
+      "loss": 5.1299,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938798961976935e-05,
+      "loss": 5.1276,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937960367225883e-05,
+      "loss": 5.1282,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937121772474831e-05,
+      "loss": 5.1223,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936284815604152e-05,
+      "loss": 5.099,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354462208531e-05,
+      "loss": 5.1024,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934607626102048e-05,
+      "loss": 5.0889,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933769031350996e-05,
+      "loss": 5.0915,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932932074480317e-05,
+      "loss": 5.0867,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932093479729265e-05,
+      "loss": 5.076,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931254884978213e-05,
+      "loss": 5.078,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930416290227161e-05,
+      "loss": 5.0683,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929577695476109e-05,
+      "loss": 5.0516,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0508,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279021438543785e-05,
+      "loss": 5.0557,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0489,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0365,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0226,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0284,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0152,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922872213228439e-05,
+      "loss": 5.0247,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922033618477387e-05,
+      "loss": 5.0043,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921195023726335e-05,
+      "loss": 4.9997,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0168,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 4.9837,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918680877353552e-05,
+      "loss": 4.9776,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917842282602501e-05,
+      "loss": 4.9822,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917003687851449e-05,
+      "loss": 4.9802,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9811,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9687,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.963,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913650946727614e-05,
+      "loss": 4.9546,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912812351976562e-05,
+      "loss": 4.9551,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911973757225509e-05,
+      "loss": 4.9382,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.9392,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910298205603779e-05,
+      "loss": 4.9423,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909459610852727e-05,
+      "loss": 4.9314,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908621016101675e-05,
+      "loss": 4.9248,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907782421350623e-05,
+      "loss": 4.92,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906943826599571e-05,
+      "loss": 4.9109,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906105231848519e-05,
+      "loss": 4.9207,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052682749778396e-05,
+      "loss": 4.9065,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044296802267876e-05,
+      "loss": 4.9033,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9035910854757356e-05,
+      "loss": 4.9045,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027524907246836e-05,
+      "loss": 4.8985,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.8923,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.8907,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.8915,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8993997496008485e-05,
+      "loss": 4.8907,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.8697,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.8797,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.8704,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896047008477014e-05,
+      "loss": 4.8727,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895208413725962e-05,
+      "loss": 4.8536,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89436981897491e-05,
+      "loss": 4.8626,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893531224223858e-05,
+      "loss": 4.8477,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892694267353179e-05,
+      "loss": 4.8542,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891855672602127e-05,
+      "loss": 4.831,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891017077851075e-05,
+      "loss": 4.8561,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890178483100023e-05,
+      "loss": 4.8589,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8391,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8271,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766433672724e-05,
+      "loss": 4.8383,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886825741976188e-05,
+      "loss": 4.831,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859871472251365e-05,
+      "loss": 4.829,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851485524740845e-05,
+      "loss": 4.8236,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843099577230325e-05,
+      "loss": 4.8268,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834713629719805e-05,
+      "loss": 4.8167,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8112,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.7965,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.808,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8048,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879281664977466e-05,
+      "loss": 4.7955,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878443070226414e-05,
+      "loss": 4.8082,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877604475475362e-05,
+      "loss": 4.7873,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87676588072431e-05,
+      "loss": 4.7961,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.7969,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.7865,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.7972731590271,
+      "eval_runtime": 313.984,
+      "eval_samples_per_second": 1215.32,
+      "eval_steps_per_second": 37.98,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.7745,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.7801,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872574544849423e-05,
+      "loss": 4.7839,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871735950098371e-05,
+      "loss": 4.7728,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870897355347319e-05,
+      "loss": 4.7752,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870058760596267e-05,
+      "loss": 4.758,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7629,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7537,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7592,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.75,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.7557,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.7575,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7424,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7396,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7257,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7365,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7299,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7238,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.721,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.7489,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.7194,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.7224,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7253,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7146,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.7129,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.7014,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.7086,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851617865474989e-05,
+      "loss": 4.7065,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.6927,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7015,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849103719102206e-05,
+      "loss": 4.6906,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848265124351154e-05,
+      "loss": 4.7008,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847426529600102e-05,
+      "loss": 4.6963,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.6844,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.6944,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.7025,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.6713,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.6818,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.6775,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.6692,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.6715,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839880814721007e-05,
+      "loss": 4.6795,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.6717,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.6668,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837368306228597e-05,
+      "loss": 4.6697,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836529711477545e-05,
+      "loss": 4.6626,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.6672,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6494,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.6453,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.6637,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832338375602659e-05,
+      "loss": 4.6606,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831499780851606e-05,
+      "loss": 4.6499,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830661186100554e-05,
+      "loss": 4.6413,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829822591349502e-05,
+      "loss": 4.6504,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6328,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6605,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.6355,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.6493,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.6324,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.6271,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.6301,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82311547122146e-05,
+      "loss": 4.6351,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6169,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214415574801015e-05,
+      "loss": 4.6218,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206029627290495e-05,
+      "loss": 4.6346,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197643679779975e-05,
+      "loss": 4.6163,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189257732269455e-05,
+      "loss": 4.6085,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6067,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.6063,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.6283,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.6201,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6057,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.61,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6136,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6115,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.5961,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.6081,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.5949,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808867549855441e-05,
+      "loss": 4.6039,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.5949,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.5951,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.5975,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.5985,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804676213980554e-05,
+      "loss": 4.5808,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803837619229502e-05,
+      "loss": 4.5902,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.5887,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.5925,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.586,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.5727,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799646283354615e-05,
+      "loss": 4.583,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798807688603563e-05,
+      "loss": 4.5752,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5876,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5671,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5717,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.5835,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5721,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5541,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5655,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5702,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5724,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5625,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895896978634836e-05,
+      "loss": 4.5601,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887511031124316e-05,
+      "loss": 4.5552,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.5603,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.5413,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5502,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5564,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5536,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5373,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782884215615814e-05,
+      "loss": 4.5474,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.5357,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.551,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.538,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.5305,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5478,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5408,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.536,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5379,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5373,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5417,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5177,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5395,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5267,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5308,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5224,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.527,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.5165,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.525,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5027,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5365,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5285,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5216,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5085,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5218,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5228,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610856657295813e-05,
+      "loss": 4.5138,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602470709785293e-05,
+      "loss": 4.5085,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594084762274773e-05,
+      "loss": 4.5184,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5118,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5188,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.4935,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.51,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75521714035259e-05,
+      "loss": 4.5066,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543785456015375e-05,
+      "loss": 4.4998,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.5152,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.4959,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.504,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5012,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750187209726651e-05,
+      "loss": 4.502,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.516730785369873,
+      "eval_runtime": 330.6418,
+      "eval_samples_per_second": 1154.092,
+      "eval_steps_per_second": 36.066,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749348614975599e-05,
+      "loss": 4.4896,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748510020224547e-05,
+      "loss": 4.4968,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.503,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746832830722443e-05,
+      "loss": 4.497,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745994235971391e-05,
+      "loss": 4.4973,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.4807,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.489,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.4813,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.4915,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.4826,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.4921,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.492,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4751,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.4806,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.4626,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736769693709819e-05,
+      "loss": 4.4756,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735931098958767e-05,
+      "loss": 4.4753,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4683,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.468,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.4929,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4717,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.4712,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4762,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.473,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4663,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4575,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4669,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4624,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4548,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4644,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4542,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4638,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4596,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4498,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4623,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4746,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719165755459219e-05,
+      "loss": 4.4416,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718327160708167e-05,
+      "loss": 4.4574,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717488565957115e-05,
+      "loss": 4.4451,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716649971206063e-05,
+      "loss": 4.4473,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4441,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.459,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714135824833281e-05,
+      "loss": 4.4508,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4475,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4489,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.4411,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4533,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4337,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4319,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4475,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707428704705237e-05,
+      "loss": 4.4452,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.4367,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705751515203134e-05,
+      "loss": 4.4327,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049145583324546e-05,
+      "loss": 4.441,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040759635814026e-05,
+      "loss": 4.4234,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032373688303506e-05,
+      "loss": 4.4468,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7023987740792986e-05,
+      "loss": 4.4291,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015601793282466e-05,
+      "loss": 4.4421,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4341,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4246,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4266,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4346,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4147,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965318865826804e-05,
+      "loss": 4.4239,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695693291831629e-05,
+      "loss": 4.4323,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694854697080577e-05,
+      "loss": 4.4228,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4081,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4107,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4085,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4356,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4241,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4143,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4232,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688149214832907e-05,
+      "loss": 4.4203,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687310620081855e-05,
+      "loss": 4.4238,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686472025330803e-05,
+      "loss": 4.4063,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.4204,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4128,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.4141,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.4087,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822806894559164e-05,
+      "loss": 4.4122,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814420947048644e-05,
+      "loss": 4.4114,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.416,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797665430831333e-05,
+      "loss": 4.4011,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678927948332081e-05,
+      "loss": 4.407,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6780893535810287e-05,
+      "loss": 4.405,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772507588299766e-05,
+      "loss": 4.4108,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764121640789246e-05,
+      "loss": 4.4101,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755735693278726e-05,
+      "loss": 4.3914,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.405,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673898017706142e-05,
+      "loss": 4.4005,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4095,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.3884,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.3997,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.4112,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.3986,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.3808,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668029492329151e-05,
+      "loss": 4.3931,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667190897578099e-05,
+      "loss": 4.4011,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635394070742e-05,
+      "loss": 4.4009,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665515345956368e-05,
+      "loss": 4.3951,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664676751205316e-05,
+      "loss": 4.3869,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663838156454265e-05,
+      "loss": 4.3858,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.3924,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621626048325336e-05,
+      "loss": 4.3776,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613240100814816e-05,
+      "loss": 4.3843,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3903,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.3872,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.37,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579729068380194e-05,
+      "loss": 4.3873,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571343120869674e-05,
+      "loss": 4.372,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.3851,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554571225848634e-05,
+      "loss": 4.3793,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.3704,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65377993308276e-05,
+      "loss": 4.3859,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.3835,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.3729,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.3792,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.3796,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649588597207873e-05,
+      "loss": 4.38,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648750002456821e-05,
+      "loss": 4.3655,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.3806,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647072812954717e-05,
+      "loss": 4.3719,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646234218203665e-05,
+      "loss": 4.374,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.3696,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3774,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.358,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.3723,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420445202091514e-05,
+      "loss": 4.3515,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412059254580994e-05,
+      "loss": 4.3823,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403673307070474e-05,
+      "loss": 4.3735,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639528735955995e-05,
+      "loss": 4.3706,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.3576,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637853184334264e-05,
+      "loss": 4.37,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.3752,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3628,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.3619,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6344988053300556e-05,
+      "loss": 4.3697,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336602105790036e-05,
+      "loss": 4.364,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3716,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3515,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3638,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3611,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3539,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628630279953065e-05,
+      "loss": 4.3727,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627791685202013e-05,
+      "loss": 4.3511,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626953090450961e-05,
+      "loss": 4.3601,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.3617,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.3567,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.382836818695068,
+      "eval_runtime": 295.0935,
+      "eval_samples_per_second": 1293.119,
+      "eval_steps_per_second": 40.411,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3548,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3541,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3634,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3578,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.3558,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202508839640374e-05,
+      "loss": 4.3448,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3518,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3356,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.3605,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3454,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.355,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.3568,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61438399646742e-05,
+      "loss": 4.3438,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613545401716368e-05,
+      "loss": 4.3388,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3354,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3352,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.3449,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.3346,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.335,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.3593,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607675238459004e-05,
+      "loss": 4.3418,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606836643707952e-05,
+      "loss": 4.3407,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059980489569e-05,
+      "loss": 4.3459,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605159454205848e-05,
+      "loss": 4.3421,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3369,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.3267,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3415,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601806713082013e-05,
+      "loss": 4.3267,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009697562113344e-05,
+      "loss": 4.3301,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001311614602824e-05,
+      "loss": 4.3302,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5992925667092303e-05,
+      "loss": 4.3265,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984539719581783e-05,
+      "loss": 4.3403,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976153772071257e-05,
+      "loss": 4.3314,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967767824560737e-05,
+      "loss": 4.323,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595941463465768e-05,
+      "loss": 4.3377,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595102868714716e-05,
+      "loss": 4.3438,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3223,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3323,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3163,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3329,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3169,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3364,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589234343337726e-05,
+      "loss": 4.3283,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588395748586673e-05,
+      "loss": 4.3251,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587557153835621e-05,
+      "loss": 4.3289,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586718559084569e-05,
+      "loss": 4.3189,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3314,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3147,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3103,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3272,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3214,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816902663390035e-05,
+      "loss": 4.3226,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3105,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3209,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3068,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783375252151684e-05,
+      "loss": 4.3266,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774989304641164e-05,
+      "loss": 4.3132,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766603357130644e-05,
+      "loss": 4.3274,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758217409620124e-05,
+      "loss": 4.3192,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574984784091333e-05,
+      "loss": 4.3112,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574146189340281e-05,
+      "loss": 4.3069,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573307594589229e-05,
+      "loss": 4.3183,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572468999838177e-05,
+      "loss": 4.3013,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571630405087125e-05,
+      "loss": 4.3078,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570791810336074e-05,
+      "loss": 4.315,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569953215585022e-05,
+      "loss": 4.3137,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56911462083397e-05,
+      "loss": 4.2935,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568276026082918e-05,
+      "loss": 4.2989,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.2933,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3245,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565761879710135e-05,
+      "loss": 4.3105,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564923284959083e-05,
+      "loss": 4.3051,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3107,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.306,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3186,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561570543835248e-05,
+      "loss": 4.2918,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560731949084196e-05,
+      "loss": 4.3132,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559893354333144e-05,
+      "loss": 4.3016,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559054759582092e-05,
+      "loss": 4.3053,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55821616483104e-05,
+      "loss": 4.3041,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.2982,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3073,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3046,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554865061587578e-05,
+      "loss": 4.2931,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554026466836526e-05,
+      "loss": 4.3023,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553187872085474e-05,
+      "loss": 4.2928,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523492773344215e-05,
+      "loss": 4.3088,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515106825833695e-05,
+      "loss": 4.3016,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.2864,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549835130961639e-05,
+      "loss": 4.3002,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548996536210587e-05,
+      "loss": 4.2906,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548157941459535e-05,
+      "loss": 4.3002,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547319346708483e-05,
+      "loss": 4.2865,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546480751957431e-05,
+      "loss": 4.2951,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545642157206379e-05,
+      "loss": 4.3047,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544803562455327e-05,
+      "loss": 4.2903,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.2831,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543128010833596e-05,
+      "loss": 4.2898,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542289416082544e-05,
+      "loss": 4.2913,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541452459211865e-05,
+      "loss": 4.3041,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540613864460813e-05,
+      "loss": 4.2893,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539775269709761e-05,
+      "loss": 4.2866,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389366749587095e-05,
+      "loss": 4.2833,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380997180880304e-05,
+      "loss": 4.2919,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372611233369784e-05,
+      "loss": 4.2718,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364225285859264e-05,
+      "loss": 4.2849,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355839338348744e-05,
+      "loss": 4.2868,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347453390838224e-05,
+      "loss": 4.2896,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339067443327704e-05,
+      "loss": 4.2652,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533069787462091e-05,
+      "loss": 4.2863,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532231192711039e-05,
+      "loss": 4.2789,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531392597959987e-05,
+      "loss": 4.2758,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530554003208935e-05,
+      "loss": 4.2835,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529715408457883e-05,
+      "loss": 4.2683,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528876813706831e-05,
+      "loss": 4.2902,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528039856836153e-05,
+      "loss": 4.2837,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527201262085101e-05,
+      "loss": 4.2751,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526362667334049e-05,
+      "loss": 4.2829,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525524072582997e-05,
+      "loss": 4.2771,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524685477831945e-05,
+      "loss": 4.2843,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523848520961266e-05,
+      "loss": 4.262,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523009926210214e-05,
+      "loss": 4.2889,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522171331459162e-05,
+      "loss": 4.2735,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52133273670811e-05,
+      "loss": 4.2751,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520494141957057e-05,
+      "loss": 4.2753,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519657185086379e-05,
+      "loss": 4.284,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518818590335327e-05,
+      "loss": 4.2664,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517979995584275e-05,
+      "loss": 4.2724,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517141400833223e-05,
+      "loss": 4.2608,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516302806082171e-05,
+      "loss": 4.2844,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154642113311187e-05,
+      "loss": 4.2776,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514628892340813e-05,
+      "loss": 4.2766,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513790297589761e-05,
+      "loss": 4.2641,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512951702838709e-05,
+      "loss": 4.2759,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512113108087657e-05,
+      "loss": 4.2798,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112745133366045e-05,
+      "loss": 4.2705,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510437556465926e-05,
+      "loss": 4.2673,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509598961714874e-05,
+      "loss": 4.2763,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508760366963822e-05,
+      "loss": 4.2746,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50792177221277e-05,
+      "loss": 4.276,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507083177461718e-05,
+      "loss": 4.2598,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506244582710666e-05,
+      "loss": 4.2719,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505405987959614e-05,
+      "loss": 4.2668,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504567393208562e-05,
+      "loss": 4.2665,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50372879845751e-05,
+      "loss": 4.275,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502890203706458e-05,
+      "loss": 4.2621,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502051608955406e-05,
+      "loss": 4.2682,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501214652084727e-05,
+      "loss": 4.2657,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500376057333675e-05,
+      "loss": 4.272,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.300319671630859,
+      "eval_runtime": 295.0933,
+      "eval_samples_per_second": 1293.12,
+      "eval_steps_per_second": 40.411,
+      "step": 305280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.499537462582623e-05,
+      "loss": 4.2631,
+      "step": 305664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.498698867831571e-05,
+      "loss": 4.2656,
+      "step": 306176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.497860273080519e-05,
+      "loss": 4.2722,
+      "step": 306688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.497021678329467e-05,
+      "loss": 4.2653,
+      "step": 307200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.496183083578415e-05,
+      "loss": 4.2703,
+      "step": 307712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4953461267077365e-05,
+      "loss": 4.254,
+      "step": 308224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4945075319566845e-05,
+      "loss": 4.2636,
+      "step": 308736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4936705750860054e-05,
+      "loss": 4.2497,
+      "step": 309248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4928319803349534e-05,
+      "loss": 4.2698,
+      "step": 309760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4919933855839014e-05,
+      "loss": 4.2629,
+      "step": 310272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.491156428713222e-05,
+      "loss": 4.2641,
+      "step": 310784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.49031783396217e-05,
+      "loss": 4.2639,
+      "step": 311296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.489479239211118e-05,
+      "loss": 4.2569,
+      "step": 311808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.488640644460066e-05,
+      "loss": 4.2517,
+      "step": 312320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.487802049709014e-05,
+      "loss": 4.2543,
+      "step": 312832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.486963454957962e-05,
+      "loss": 4.2433,
+      "step": 313344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.48612486020691e-05,
+      "loss": 4.2617,
+      "step": 313856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.485286265455859e-05,
+      "loss": 4.2473,
+      "step": 314368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.48444930858518e-05,
+      "loss": 4.2496,
+      "step": 314880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.483610713834128e-05,
+      "loss": 4.2788,
+      "step": 315392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.482772119083076e-05,
+      "loss": 4.2529,
+      "step": 315904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.481933524332024e-05,
+      "loss": 4.2565,
+      "step": 316416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.481094929580971e-05,
+      "loss": 4.2599,
+      "step": 316928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.480256334829919e-05,
+      "loss": 4.2567,
+      "step": 317440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.479419377959241e-05,
+      "loss": 4.2533,
+      "step": 317952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.478580783208188e-05,
+      "loss": 4.2449,
+      "step": 318464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.477742188457136e-05,
+      "loss": 4.2576,
+      "step": 318976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.476903593706084e-05,
+      "loss": 4.2365,
+      "step": 319488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.476064998955033e-05,
+      "loss": 4.2471,
+      "step": 320000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.475226404203981e-05,
+      "loss": 4.2435,
+      "step": 320512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.474387809452929e-05,
+      "loss": 4.2441,
+      "step": 321024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.473549214701877e-05,
+      "loss": 4.2557,
+      "step": 321536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.472710619950825e-05,
+      "loss": 4.2503,
+      "step": 322048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4718736630801456e-05,
+      "loss": 4.2447,
+      "step": 322560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4710350683290936e-05,
+      "loss": 4.2536,
+      "step": 323072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4701964735780416e-05,
+      "loss": 4.2575,
+      "step": 323584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4693578788269896e-05,
+      "loss": 4.2409,
+      "step": 324096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4685192840759376e-05,
+      "loss": 4.2499,
+      "step": 324608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4676806893248856e-05,
+      "loss": 4.235,
+      "step": 325120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4668437324542065e-05,
+      "loss": 4.2522,
+      "step": 325632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.466006775583528e-05,
+      "loss": 4.2316,
+      "step": 326144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.465168180832476e-05,
+      "loss": 4.256,
+      "step": 326656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.464329586081424e-05,
+      "loss": 4.247,
+      "step": 327168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.463490991330372e-05,
+      "loss": 4.2409,
+      "step": 327680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.46265239657932e-05,
+      "loss": 4.2473,
+      "step": 328192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.461813801828268e-05,
+      "loss": 4.2392,
+      "step": 328704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.460975207077216e-05,
+      "loss": 4.2491,
+      "step": 329216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.460136612326164e-05,
+      "loss": 4.2357,
+      "step": 329728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.459299655455485e-05,
+      "loss": 4.2319,
+      "step": 330240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.458461060704433e-05,
+      "loss": 4.2504,
+      "step": 330752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.457622465953381e-05,
+      "loss": 4.2416,
+      "step": 331264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.456783871202329e-05,
+      "loss": 4.2422,
+      "step": 331776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.45594691433165e-05,
+      "loss": 4.2334,
+      "step": 332288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.455108319580598e-05,
+      "loss": 4.2391,
+      "step": 332800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4542713627099194e-05,
+      "loss": 4.2323,
+      "step": 333312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4534327679588674e-05,
+      "loss": 4.2437,
+      "step": 333824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4525941732078154e-05,
+      "loss": 4.2337,
+      "step": 334336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4517555784567634e-05,
+      "loss": 4.2494,
+      "step": 334848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4509169837057114e-05,
+      "loss": 4.2401,
+      "step": 335360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4500783889546594e-05,
+      "loss": 4.2329,
+      "step": 335872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.449239794203607e-05,
+      "loss": 4.2313,
+      "step": 336384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.448401199452555e-05,
+      "loss": 4.2385,
+      "step": 336896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.447562604701503e-05,
+      "loss": 4.2227,
+      "step": 337408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.446724009950451e-05,
+      "loss": 4.2335,
+      "step": 337920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.445885415199399e-05,
+      "loss": 4.2352,
+      "step": 338432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.44504845832872e-05,
+      "loss": 4.2392,
+      "step": 338944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.444209863577668e-05,
+      "loss": 4.2169,
+      "step": 339456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.443371268826616e-05,
+      "loss": 4.2245,
+      "step": 339968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.442532674075564e-05,
+      "loss": 4.2171,
+      "step": 340480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.441695717204885e-05,
+      "loss": 4.2444,
+      "step": 340992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.440857122453833e-05,
+      "loss": 4.2327,
+      "step": 341504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.440018527702781e-05,
+      "loss": 4.2306,
+      "step": 342016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.439179932951729e-05,
+      "loss": 4.2338,
+      "step": 342528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.438341338200677e-05,
+      "loss": 4.2308,
+      "step": 343040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.437502743449625e-05,
+      "loss": 4.2472,
+      "step": 343552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2156,
+      "step": 344064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.435825553947521e-05,
+      "loss": 4.2348,
+      "step": 344576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.434988597076842e-05,
+      "loss": 4.2313,
+      "step": 345088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.43415000232579e-05,
+      "loss": 4.2281,
+      "step": 345600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.433311407574738e-05,
+      "loss": 4.2299,
+      "step": 346112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.432472812823687e-05,
+      "loss": 4.2256,
+      "step": 346624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.431634218072635e-05,
+      "loss": 4.2301,
+      "step": 347136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.430795623321583e-05,
+      "loss": 4.2323,
+      "step": 347648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2173,
+      "step": 348160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4291200716998516e-05,
+      "loss": 4.231,
+      "step": 348672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.221,
+      "step": 349184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2332,
+      "step": 349696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2256,
+      "step": 350208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.425765692695643e-05,
+      "loss": 4.2134,
+      "step": 350720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.424927097944591e-05,
+      "loss": 4.2256,
+      "step": 351232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.424088503193539e-05,
+      "loss": 4.2205,
+      "step": 351744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4232531842032334e-05,
+      "loss": 4.2301,
+      "step": 352256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.422414589452182e-05,
+      "loss": 4.2097,
+      "step": 352768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.42157599470113e-05,
+      "loss": 4.2246,
+      "step": 353280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.420737399950078e-05,
+      "loss": 4.238,
+      "step": 353792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.419900443079399e-05,
+      "loss": 4.2181,
+      "step": 354304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.419061848328347e-05,
+      "loss": 4.206,
+      "step": 354816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.418223253577295e-05,
+      "loss": 4.2173,
+      "step": 355328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.417384658826243e-05,
+      "loss": 4.2246,
+      "step": 355840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.229,
+      "step": 356352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2224,
+      "step": 356864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.41487051245346e-05,
+      "loss": 4.2146,
+      "step": 357376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.414031917702408e-05,
+      "loss": 4.213,
+      "step": 357888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2222,
+      "step": 358400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.1992,
+      "step": 358912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.411516133449252e-05,
+      "loss": 4.2134,
+      "step": 359424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4106791765785735e-05,
+      "loss": 4.2177,
+      "step": 359936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.409840581827521e-05,
+      "loss": 4.2179,
+      "step": 360448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.1965,
+      "step": 360960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2153,
+      "step": 361472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2078,
+      "step": 361984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.2099,
+      "step": 362496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2115,
+      "step": 363008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.2027,
+      "step": 363520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.40397205645053e-05,
+      "loss": 4.216,
+      "step": 364032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.403135099579851e-05,
+      "loss": 4.2199,
+      "step": 364544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.402296504828799e-05,
+      "loss": 4.2059,
+      "step": 365056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.401457910077747e-05,
+      "loss": 4.2157,
+      "step": 365568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.400619315326695e-05,
+      "loss": 4.206,
+      "step": 366080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.399780720575643e-05,
+      "loss": 4.2184,
+      "step": 366592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.1968,
+      "step": 367104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.2212,
+      "step": 367616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.39726657420286e-05,
+      "loss": 4.2003,
+      "step": 368128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.396427979451808e-05,
+      "loss": 4.2101,
+      "step": 368640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.395589384700756e-05,
+      "loss": 4.2016,
+      "step": 369152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.394750789949704e-05,
+      "loss": 4.2171,
+      "step": 369664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.393912195198652e-05,
+      "loss": 4.2039,
+      "step": 370176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3930736004476e-05,
+      "loss": 4.2054,
+      "step": 370688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.392235005696548e-05,
+      "loss": 4.1907,
+      "step": 371200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.391396410945496e-05,
+      "loss": 4.222,
+      "step": 371712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3905610919551906e-05,
+      "loss": 4.2067,
+      "step": 372224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.21,
+      "step": 372736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3888839024530866e-05,
+      "loss": 4.1981,
+      "step": 373248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2111,
+      "step": 373760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2153,
+      "step": 374272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2035,
+      "step": 374784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3855311613292515e-05,
+      "loss": 4.2012,
+      "step": 375296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3846925665781995e-05,
+      "loss": 4.2105,
+      "step": 375808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3838556097075204e-05,
+      "loss": 4.2107,
+      "step": 376320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3830170149564684e-05,
+      "loss": 4.2112,
+      "step": 376832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3821784202054164e-05,
+      "loss": 4.1916,
+      "step": 377344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3813398254543644e-05,
+      "loss": 4.2065,
+      "step": 377856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.380501230703313e-05,
+      "loss": 4.1991,
+      "step": 378368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2032,
+      "step": 378880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2097,
+      "step": 379392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.1925,
+      "step": 379904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2099,
+      "step": 380416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.376309894828426e-05,
+      "loss": 4.1988,
+      "step": 380928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2106,
+      "step": 381440
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.244948863983154,
+      "eval_runtime": 295.8412,
+      "eval_samples_per_second": 1289.851,
+      "eval_steps_per_second": 40.309,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.1911,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.204,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2054,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371278326322113e-05,
+      "loss": 4.204,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370439731571061e-05,
+      "loss": 4.1933,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36960113682001e-05,
+      "loss": 4.1973,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368762542068958e-05,
+      "loss": 4.187,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367923947317906e-05,
+      "loss": 4.2052,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367085352566854e-05,
+      "loss": 4.2027,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366246757815802e-05,
+      "loss": 4.1991,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36540816306475e-05,
+      "loss": 4.2022,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364571206194071e-05,
+      "loss": 4.1914,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363732611443019e-05,
+      "loss": 4.1907,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362894016691967e-05,
+      "loss": 4.191,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362055421940915e-05,
+      "loss": 4.181,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361216827189863e-05,
+      "loss": 4.1959,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36037823243881e-05,
+      "loss": 4.184,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359541275568132e-05,
+      "loss": 4.188,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35870268081708e-05,
+      "loss": 4.2224,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3578640860660284e-05,
+      "loss": 4.1881,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357025491314976e-05,
+      "loss": 4.1961,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356186896563924e-05,
+      "loss": 4.1956,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355348301812872e-05,
+      "loss": 4.1948,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35450970706182e-05,
+      "loss": 4.1884,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3536711123107677e-05,
+      "loss": 4.1865,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528325175597157e-05,
+      "loss": 4.1946,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3519939228086636e-05,
+      "loss": 4.1773,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511553280576116e-05,
+      "loss": 4.1855,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503167333065596e-05,
+      "loss": 4.182,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3494797764358805e-05,
+      "loss": 4.1864,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486411816848285e-05,
+      "loss": 4.1916,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3478025869337765e-05,
+      "loss": 4.1914,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346965630063098e-05,
+      "loss": 4.1853,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346127035312046e-05,
+      "loss": 4.1925,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345288440560994e-05,
+      "loss": 4.1951,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344449845809942e-05,
+      "loss": 4.1817,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34361125105889e-05,
+      "loss": 4.1939,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342774294188211e-05,
+      "loss": 4.173,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341935699437159e-05,
+      "loss": 4.1971,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341097104686107e-05,
+      "loss": 4.1685,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340258509935055e-05,
+      "loss": 4.201,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339419915184003e-05,
+      "loss": 4.1875,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338581320432951e-05,
+      "loss": 4.1851,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337744363562272e-05,
+      "loss": 4.1889,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33690576881122e-05,
+      "loss": 4.1825,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360671740601686e-05,
+      "loss": 4.1874,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352285793091166e-05,
+      "loss": 4.1747,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343916224384375e-05,
+      "loss": 4.1662,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3335530276873855e-05,
+      "loss": 4.1958,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327144329363335e-05,
+      "loss": 4.1828,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3318758381852815e-05,
+      "loss": 4.1903,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331037243434229e-05,
+      "loss": 4.1735,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3302002865635504e-05,
+      "loss": 4.1803,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293616918124984e-05,
+      "loss": 4.1732,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285230970614464e-05,
+      "loss": 4.1847,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327684502310394e-05,
+      "loss": 4.1747,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268459075593424e-05,
+      "loss": 4.1907,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260073128082904e-05,
+      "loss": 4.1868,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251687180572383e-05,
+      "loss": 4.1714,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243301233061863e-05,
+      "loss": 4.1747,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323493166435507e-05,
+      "loss": 4.1832,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322654571684455e-05,
+      "loss": 4.1663,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321815976933403e-05,
+      "loss": 4.1741,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320977382182351e-05,
+      "loss": 4.179,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320140425311672e-05,
+      "loss": 4.179,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31930183056062e-05,
+      "loss": 4.1619,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318463235809568e-05,
+      "loss": 4.168,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317624641058516e-05,
+      "loss": 4.1601,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316786046307464e-05,
+      "loss": 4.1865,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315947451556412e-05,
+      "loss": 4.1747,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315108856805361e-05,
+      "loss": 4.1754,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314270262054309e-05,
+      "loss": 4.1804,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31343330518363e-05,
+      "loss": 4.1716,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3125963483129506e-05,
+      "loss": 4.1913,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117577535618986e-05,
+      "loss": 4.1621,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109191588108466e-05,
+      "loss": 4.1766,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3100805640597946e-05,
+      "loss": 4.1749,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092419693087426e-05,
+      "loss": 4.1716,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084033745576906e-05,
+      "loss": 4.1769,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3075647798066386e-05,
+      "loss": 4.1705,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067278229359595e-05,
+      "loss": 4.1726,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058892281849075e-05,
+      "loss": 4.1806,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305050633433856e-05,
+      "loss": 4.1605,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304212038682804e-05,
+      "loss": 4.1788,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303373443931752e-05,
+      "loss": 4.1645,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025348491807e-05,
+      "loss": 4.1791,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016962544296475e-05,
+      "loss": 4.1662,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008576596785955e-05,
+      "loss": 4.1637,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300020702807917e-05,
+      "loss": 4.1699,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299182108056865e-05,
+      "loss": 4.1699,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298345151186186e-05,
+      "loss": 4.1715,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297506556435134e-05,
+      "loss": 4.1565,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296667961684082e-05,
+      "loss": 4.1728,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29582936693303e-05,
+      "loss": 4.1807,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294990772181978e-05,
+      "loss": 4.1685,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294152177430926e-05,
+      "loss": 4.1493,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293313582679874e-05,
+      "loss": 4.1656,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292474987928822e-05,
+      "loss": 4.1687,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291638031058143e-05,
+      "loss": 4.1729,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290799436307091e-05,
+      "loss": 4.172,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289960841556039e-05,
+      "loss": 4.1621,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289122246804987e-05,
+      "loss": 4.1585,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288285289934308e-05,
+      "loss": 4.17,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287446695183256e-05,
+      "loss": 4.1492,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286608100432204e-05,
+      "loss": 4.16,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285771143561525e-05,
+      "loss": 4.1604,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284932548810473e-05,
+      "loss": 4.1678,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284093954059421e-05,
+      "loss": 4.1469,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283255359308369e-05,
+      "loss": 4.1592,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28241840243769e-05,
+      "loss": 4.1591,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281579807686638e-05,
+      "loss": 4.1564,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280741212935586e-05,
+      "loss": 4.1605,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279902618184534e-05,
+      "loss": 4.1481,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279064023433482e-05,
+      "loss": 4.1674,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278227066562803e-05,
+      "loss": 4.1657,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277388471811751e-05,
+      "loss": 4.1525,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276549877060699e-05,
+      "loss": 4.1664,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275711282309647e-05,
+      "loss": 4.1525,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274872687558595e-05,
+      "loss": 4.1697,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274034092807543e-05,
+      "loss": 4.1522,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273197135936865e-05,
+      "loss": 4.1655,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2723585411858127e-05,
+      "loss": 4.1486,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715199464347607e-05,
+      "loss": 4.1589,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2706813516837087e-05,
+      "loss": 4.1538,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698427569326566e-05,
+      "loss": 4.1626,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690041621816046e-05,
+      "loss": 4.1481,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2681655674305526e-05,
+      "loss": 4.1591,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673269726795006e-05,
+      "loss": 4.1368,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664900158088215e-05,
+      "loss": 4.1727,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656514210577695e-05,
+      "loss": 4.1533,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648144641870904e-05,
+      "loss": 4.1594,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639758694360384e-05,
+      "loss": 4.1473,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263137274684987e-05,
+      "loss": 4.1633,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262298679933935e-05,
+      "loss": 4.1626,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261460085182883e-05,
+      "loss": 4.1531,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260621490431831e-05,
+      "loss": 4.1537,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597828956807784e-05,
+      "loss": 4.1599,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1597,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258107344059048e-05,
+      "loss": 4.1683,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257270387188369e-05,
+      "loss": 4.1376,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256431792437317e-05,
+      "loss": 4.1595,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255593197686265e-05,
+      "loss": 4.1461,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254754602935213e-05,
+      "loss": 4.1572,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253916008184161e-05,
+      "loss": 4.1567,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253077413433109e-05,
+      "loss": 4.1401,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522404565624305e-05,
+      "loss": 4.1617,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514018618113785e-05,
+      "loss": 4.1536,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250563267060326e-05,
+      "loss": 4.1566,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.204551696777344,
+      "eval_runtime": 319.3114,
+      "eval_samples_per_second": 1195.043,
+      "eval_steps_per_second": 37.346,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249724672309274e-05,
+      "loss": 4.1493,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248886077558222e-05,
+      "loss": 4.15,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24804748280717e-05,
+      "loss": 4.1583,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247208888056118e-05,
+      "loss": 4.1519,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246370293305066e-05,
+      "loss": 4.1555,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245533336434387e-05,
+      "loss": 4.1497,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244694741683335e-05,
+      "loss": 4.1434,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243856146932283e-05,
+      "loss": 4.1434,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243017552181231e-05,
+      "loss": 4.1532,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242180595310552e-05,
+      "loss": 4.157,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413420005595e-05,
+      "loss": 4.1495,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240505043688821e-05,
+      "loss": 4.1549,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239668086818143e-05,
+      "loss": 4.14,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238829492067091e-05,
+      "loss": 4.1456,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237990897316038e-05,
+      "loss": 4.1444,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237152302564986e-05,
+      "loss": 4.1327,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236313707813934e-05,
+      "loss": 4.1489,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235475113062882e-05,
+      "loss": 4.1339,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23463651831183e-05,
+      "loss": 4.141,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233797923560778e-05,
+      "loss": 4.1745,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232959328809726e-05,
+      "loss": 4.1419,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.1455,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.148,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2304451824369436e-05,
+      "loss": 4.1551,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1338,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2287679929348396e-05,
+      "loss": 4.1417,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2279293981837876e-05,
+      "loss": 4.1453,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2270908034327356e-05,
+      "loss": 4.1345,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2262538465620565e-05,
+      "loss": 4.1397,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2254152518110045e-05,
+      "loss": 4.1348,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2245766570599525e-05,
+      "loss": 4.1424,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2237380623089005e-05,
+      "loss": 4.1486,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2229011054382214e-05,
+      "loss": 4.1416,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22206251068717e-05,
+      "loss": 4.1379,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221223915936118e-05,
+      "loss": 4.143,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220385321185066e-05,
+      "loss": 4.1479,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219546726434014e-05,
+      "loss": 4.135,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218708131682962e-05,
+      "loss": 4.1527,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2178695369319094e-05,
+      "loss": 4.1261,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217032580061231e-05,
+      "loss": 4.1473,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216193985310179e-05,
+      "loss": 4.1237,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215355390559126e-05,
+      "loss": 4.1536,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214516795808074e-05,
+      "loss": 4.1397,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213678201057022e-05,
+      "loss": 4.1441,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212841244186344e-05,
+      "loss": 4.142,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212002649435292e-05,
+      "loss": 4.1386,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2111656925646134e-05,
+      "loss": 4.1433,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1327,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2094885030625094e-05,
+      "loss": 4.1184,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208649908311457e-05,
+      "loss": 4.1525,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207811313560405e-05,
+      "loss": 4.1345,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206972718809353e-05,
+      "loss": 4.1483,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206134124058301e-05,
+      "loss": 4.1279,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205295529307249e-05,
+      "loss": 4.1351,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204456934556197e-05,
+      "loss": 4.1267,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1419,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.134,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.1439,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201105831312735e-05,
+      "loss": 4.1425,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200267236561683e-05,
+      "loss": 4.1264,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199428641810631e-05,
+      "loss": 4.1277,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1464,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197751452308527e-05,
+      "loss": 4.1219,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196912857557475e-05,
+      "loss": 4.1296,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196074262806423e-05,
+      "loss": 4.1325,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195235668055371e-05,
+      "loss": 4.1404,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194397073304319e-05,
+      "loss": 4.1187,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193558478553267e-05,
+      "loss": 4.1245,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192721521682588e-05,
+      "loss": 4.1219,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191882926931536e-05,
+      "loss": 4.1325,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191044332180484e-05,
+      "loss": 4.1374,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190205737429432e-05,
+      "loss": 4.1326,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18936714267838e-05,
+      "loss": 4.1352,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188528547927328e-05,
+      "loss": 4.1267,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187689953176276e-05,
+      "loss": 4.1502,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1868529963055976e-05,
+      "loss": 4.1185,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860144015545456e-05,
+      "loss": 4.1337,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185175806803493e-05,
+      "loss": 4.1314,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184337212052441e-05,
+      "loss": 4.1286,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183498617301389e-05,
+      "loss": 4.135,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18266166043071e-05,
+      "loss": 4.1289,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181823065679658e-05,
+      "loss": 4.1249,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180984470928606e-05,
+      "loss": 4.138,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180145876177554e-05,
+      "loss": 4.1194,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793072814265025e-05,
+      "loss": 4.1317,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1784703245558234e-05,
+      "loss": 4.1258,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776317298047714e-05,
+      "loss": 4.1368,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767931350537194e-05,
+      "loss": 4.1237,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1759545403026674e-05,
+      "loss": 4.1201,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1751159455516154e-05,
+      "loss": 4.1256,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1742773508005634e-05,
+      "loss": 4.1306,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173440393929884e-05,
+      "loss": 4.1298,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172601799178832e-05,
+      "loss": 4.115,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17176320442778e-05,
+      "loss": 4.1307,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170924609676728e-05,
+      "loss": 4.1378,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170086014925676e-05,
+      "loss": 4.1294,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169247420174624e-05,
+      "loss": 4.1083,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168408825423572e-05,
+      "loss": 4.1228,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16757023067252e-05,
+      "loss": 4.1222,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166731635921469e-05,
+      "loss": 4.1336,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165896316931163e-05,
+      "loss": 4.1309,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165057722180111e-05,
+      "loss": 4.1195,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164219127429059e-05,
+      "loss": 4.1168,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163380532678007e-05,
+      "loss": 4.1272,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162541937926955e-05,
+      "loss": 4.1076,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161704981056276e-05,
+      "loss": 4.1237,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160866386305224e-05,
+      "loss": 4.1136,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160027791554172e-05,
+      "loss": 4.1312,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1591891968031197e-05,
+      "loss": 4.105,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1583506020520677e-05,
+      "loss": 4.1164,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1575120073010156e-05,
+      "loss": 4.1147,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156673412549964e-05,
+      "loss": 4.1236,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155836455679285e-05,
+      "loss": 4.1167,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154997860928233e-05,
+      "loss": 4.1111,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154159266177181e-05,
+      "loss": 4.1234,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153322309306502e-05,
+      "loss": 4.1261,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15248371455545e-05,
+      "loss": 4.1134,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151645119804398e-05,
+      "loss": 4.13,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150806525053346e-05,
+      "loss": 4.1101,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1499679303022934e-05,
+      "loss": 4.1227,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491293355512414e-05,
+      "loss": 4.1136,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1482907408001894e-05,
+      "loss": 4.1225,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147452146049138e-05,
+      "loss": 4.111,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146615189178459e-05,
+      "loss": 4.118,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145776594427407e-05,
+      "loss": 4.1194,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144937999676355e-05,
+      "loss": 4.1181,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144099404925303e-05,
+      "loss": 4.1101,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143262448054624e-05,
+      "loss": 4.12,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142423853303572e-05,
+      "loss": 4.1014,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14158525855252e-05,
+      "loss": 4.129,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140746663801468e-05,
+      "loss": 4.1161,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139908069050416e-05,
+      "loss": 4.1189,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1390727500601104e-05,
+      "loss": 4.1087,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382341553090584e-05,
+      "loss": 4.1293,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1373955605580064e-05,
+      "loss": 4.1188,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1365569658069544e-05,
+      "loss": 4.115,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357183710559024e-05,
+      "loss": 4.1141,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1348797763048504e-05,
+      "loss": 4.1193,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340411815537984e-05,
+      "loss": 4.1191,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1332025868027464e-05,
+      "loss": 4.1291,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1323639920516944e-05,
+      "loss": 4.0992,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1315253973006424e-05,
+      "loss": 4.12,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1306868025495903e-05,
+      "loss": 4.1095,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1298482077985383e-05,
+      "loss": 4.1202,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129009613047486e-05,
+      "loss": 4.1142,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128171018296434e-05,
+      "loss": 4.1059,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127332423545382e-05,
+      "loss": 4.1198,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126497104555077e-05,
+      "loss": 4.117,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125658509804025e-05,
+      "loss": 4.118,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.173418045043945,
+      "eval_runtime": 291.3486,
+      "eval_samples_per_second": 1309.74,
+      "eval_steps_per_second": 40.93,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124819915052973e-05,
+      "loss": 4.1149,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123981320301921e-05,
+      "loss": 4.1101,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123142725550869e-05,
+      "loss": 4.1186,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122304130799817e-05,
+      "loss": 4.1171,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121465536048765e-05,
+      "loss": 4.1189,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120626941297712e-05,
+      "loss": 4.107,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11978834654666e-05,
+      "loss": 4.1051,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118949751795608e-05,
+      "loss": 4.1015,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118111157044556e-05,
+      "loss": 4.1163,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117272562293504e-05,
+      "loss": 4.1228,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116433967542452e-05,
+      "loss": 4.1075,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1155953727914e-05,
+      "loss": 4.1154,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114758415920722e-05,
+      "loss": 4.1061,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11391982116967e-05,
+      "loss": 4.1059,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113081226418618e-05,
+      "loss": 4.104,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112242631667566e-05,
+      "loss": 4.0996,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111404036916514e-05,
+      "loss": 4.1066,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1105654421654617e-05,
+      "loss": 4.1012,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1097268474144097e-05,
+      "loss": 4.0997,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1088898905437306e-05,
+      "loss": 4.1327,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1080512957926786e-05,
+      "loss": 4.109,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072127010416266e-05,
+      "loss": 4.1066,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1063741062905746e-05,
+      "loss": 4.1096,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1055355115395225e-05,
+      "loss": 4.1132,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1046969167884705e-05,
+      "loss": 4.098,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038583220374185e-05,
+      "loss": 4.1064,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030197272863665e-05,
+      "loss": 4.1067,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021811325353145e-05,
+      "loss": 4.0991,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101344175664636e-05,
+      "loss": 4.096,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1005055809135834e-05,
+      "loss": 4.0968,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0996669861625314e-05,
+      "loss": 4.104,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988283914114794e-05,
+      "loss": 4.1115,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0979897966604274e-05,
+      "loss": 4.1051,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0971512019093754e-05,
+      "loss": 4.1002,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963126071583234e-05,
+      "loss": 4.1069,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095475650287644e-05,
+      "loss": 4.1084,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094637055536592e-05,
+      "loss": 4.1037,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09379846078554e-05,
+      "loss": 4.1098,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092959866034489e-05,
+      "loss": 4.0915,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092121271283437e-05,
+      "loss": 4.1072,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091282676532385e-05,
+      "loss": 4.0898,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.090444081781333e-05,
+      "loss": 4.1141,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089605487030281e-05,
+      "loss": 4.1045,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088770168039975e-05,
+      "loss": 4.1056,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087931573288923e-05,
+      "loss": 4.1087,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087092978537871e-05,
+      "loss": 4.0985,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086254383786819e-05,
+      "loss": 4.1079,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085415789035767e-05,
+      "loss": 4.0964,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084578832165088e-05,
+      "loss": 4.0775,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083740237414036e-05,
+      "loss": 4.1164,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0829016426629844e-05,
+      "loss": 4.0981,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0820630479119323e-05,
+      "loss": 4.1147,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0812244531608803e-05,
+      "loss": 4.0946,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0803858584098283e-05,
+      "loss": 4.0984,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079547263658776e-05,
+      "loss": 4.0883,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078710306788097e-05,
+      "loss": 4.1057,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077871712037045e-05,
+      "loss": 4.099,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077033117285993e-05,
+      "loss": 4.1049,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076194522534941e-05,
+      "loss": 4.1117,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075355927783889e-05,
+      "loss": 4.0914,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074517333032837e-05,
+      "loss": 4.0913,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073680376162158e-05,
+      "loss": 4.1105,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072841781411106e-05,
+      "loss": 4.089,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072003186660054e-05,
+      "loss": 4.0895,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071164591909002e-05,
+      "loss": 4.0983,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07032599715795e-05,
+      "loss": 4.1026,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069489040287272e-05,
+      "loss": 4.0844,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06865044553622e-05,
+      "loss": 4.0883,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067811850785167e-05,
+      "loss": 4.0829,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066973256034115e-05,
+      "loss": 4.097,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066134661283063e-05,
+      "loss": 4.1029,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065296066532011e-05,
+      "loss": 4.0975,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064457471780959e-05,
+      "loss": 4.096,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063618877029907e-05,
+      "loss": 4.0947,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062780282278855e-05,
+      "loss": 4.1139,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0619449632885495e-05,
+      "loss": 4.083,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0611063685374975e-05,
+      "loss": 4.1019,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0602677737864455e-05,
+      "loss": 4.0965,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594291790353935e-05,
+      "loss": 4.0884,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0585905842843415e-05,
+      "loss": 4.103,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0577519895332895e-05,
+      "loss": 4.0952,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569133947822375e-05,
+      "loss": 4.0917,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0560748000311855e-05,
+      "loss": 4.1018,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0552378431605064e-05,
+      "loss": 4.0866,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0543992484094544e-05,
+      "loss": 4.0946,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0535606536584024e-05,
+      "loss": 4.092,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527220589073504e-05,
+      "loss": 4.102,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0518834641562984e-05,
+      "loss": 4.0876,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05104650728562e-05,
+      "loss": 4.0892,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050207912534568e-05,
+      "loss": 4.085,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049369317783516e-05,
+      "loss": 4.0976,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048530723032464e-05,
+      "loss": 4.0958,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047693766161785e-05,
+      "loss": 4.0808,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046855171410733e-05,
+      "loss": 4.0989,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046016576659681e-05,
+      "loss": 4.1011,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045177981908629e-05,
+      "loss": 4.0936,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044339387157577e-05,
+      "loss": 4.0742,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043502430286898e-05,
+      "loss": 4.0892,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042663835535846e-05,
+      "loss": 4.0852,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418268786651666e-05,
+      "loss": 4.1019,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040988283914115e-05,
+      "loss": 4.0979,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040149689163063e-05,
+      "loss": 4.0847,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039311094412011e-05,
+      "loss": 4.0873,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038472499660959e-05,
+      "loss": 4.0887,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037633904909907e-05,
+      "loss": 4.0791,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036795310158855e-05,
+      "loss": 4.0855,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035956715407803e-05,
+      "loss": 4.0813,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351181206567506e-05,
+      "loss": 4.0986,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0342795259056986e-05,
+      "loss": 4.0709,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0334409311546466e-05,
+      "loss": 4.0832,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0326023364035946e-05,
+      "loss": 4.0806,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0317653795329155e-05,
+      "loss": 4.0903,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0309267847818635e-05,
+      "loss": 4.0795,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030088190030812e-05,
+      "loss": 4.0784,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02924959527976e-05,
+      "loss": 4.091,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028412638409081e-05,
+      "loss": 4.0947,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027574043658029e-05,
+      "loss": 4.0773,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026735448906977e-05,
+      "loss": 4.0927,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025896854155925e-05,
+      "loss": 4.0808,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025058259404873e-05,
+      "loss": 4.0883,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024219664653821e-05,
+      "loss": 4.0816,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023382707783142e-05,
+      "loss": 4.0867,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02254411303209e-05,
+      "loss": 4.0799,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021705518281038e-05,
+      "loss": 4.0883,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020866923529986e-05,
+      "loss": 4.0868,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020028328778934e-05,
+      "loss": 4.0837,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191913719082555e-05,
+      "loss": 4.0777,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183527771572035e-05,
+      "loss": 4.0854,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175141824061515e-05,
+      "loss": 4.0682,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166755876550995e-05,
+      "loss": 4.0998,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0158386307844204e-05,
+      "loss": 4.0793,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0150000360333684e-05,
+      "loss": 4.0862,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0141614412823164e-05,
+      "loss": 4.0754,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133228465312644e-05,
+      "loss": 4.0924,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0124842517802124e-05,
+      "loss": 4.0876,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0116456570291604e-05,
+      "loss": 4.081,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010808700158481e-05,
+      "loss": 4.0844,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009970105407429e-05,
+      "loss": 4.0829,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009131510656377e-05,
+      "loss": 4.0857,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008292915905325e-05,
+      "loss": 4.1009,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007454321154274e-05,
+      "loss": 4.0639,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006615726403222e-05,
+      "loss": 4.0921,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005777131652169e-05,
+      "loss": 4.0771,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004938536901117e-05,
+      "loss": 4.0875,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004101580030439e-05,
+      "loss": 4.0792,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00326462315976e-05,
+      "loss": 4.0786,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002426028408708e-05,
+      "loss": 4.0868,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001587433657656e-05,
+      "loss": 4.086,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000750476786977e-05,
+      "loss": 4.0851,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.149842739105225,
+      "eval_runtime": 293.9403,
+      "eval_samples_per_second": 1298.192,
+      "eval_steps_per_second": 40.569,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999911882035925e-05,
+      "loss": 4.083,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999073287284873e-05,
+      "loss": 4.0773,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9982346925338207e-05,
+      "loss": 4.0861,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997396097782769e-05,
+      "loss": 4.0833,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9965575030317167e-05,
+      "loss": 4.0878,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9957189082806646e-05,
+      "loss": 4.0769,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9948803135296126e-05,
+      "loss": 4.0741,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9940417187785606e-05,
+      "loss": 4.075,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9932031240275086e-05,
+      "loss": 4.0805,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9923645292764566e-05,
+      "loss": 4.0902,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9915259345254046e-05,
+      "loss": 4.0775,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9906889776547255e-05,
+      "loss": 4.0845,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98985365866442e-05,
+      "loss": 4.08,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989015063913368e-05,
+      "loss": 4.0727,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988176469162316e-05,
+      "loss": 4.0773,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987337874411264e-05,
+      "loss": 4.0643,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986499279660212e-05,
+      "loss": 4.0767,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98566068490916e-05,
+      "loss": 4.0672,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984822090158108e-05,
+      "loss": 4.0716,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983983495407056e-05,
+      "loss": 4.0973,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983146538536377e-05,
+      "loss": 4.0837,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982307943785325e-05,
+      "loss": 4.0772,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981469349034273e-05,
+      "loss": 4.0783,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980630754283221e-05,
+      "loss": 4.0839,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979792159532169e-05,
+      "loss": 4.0612,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97895520266149e-05,
+      "loss": 4.0786,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781166079104385e-05,
+      "loss": 4.0788,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9772780131593865e-05,
+      "loss": 4.0658,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9764394184083345e-05,
+      "loss": 4.0649,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9756008236572825e-05,
+      "loss": 4.0682,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9747622289062305e-05,
+      "loss": 4.0722,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9739236341551785e-05,
+      "loss": 4.0835,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9730850394041265e-05,
+      "loss": 4.074,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9722480825334474e-05,
+      "loss": 4.0732,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9714094877823954e-05,
+      "loss": 4.0776,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705708930313434e-05,
+      "loss": 4.0772,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9697322982802913e-05,
+      "loss": 4.075,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9688937035292393e-05,
+      "loss": 4.0769,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96805674665856e-05,
+      "loss": 4.0617,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967218151907508e-05,
+      "loss": 4.0761,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.0619,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965542600285778e-05,
+      "loss": 4.0819,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964704005534726e-05,
+      "loss": 4.0767,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963865410783674e-05,
+      "loss": 4.073,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963026816032622e-05,
+      "loss": 4.0789,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96218822128157e-05,
+      "loss": 4.0699,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961351264410891e-05,
+      "loss": 4.0795,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960512669659839e-05,
+      "loss": 4.0677,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959674074908787e-05,
+      "loss": 4.0439,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958835480157735e-05,
+      "loss": 4.0846,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957996885406682e-05,
+      "loss": 4.069,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9571599285360036e-05,
+      "loss": 4.086,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956321333784952e-05,
+      "loss": 4.0652,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9554827390339e-05,
+      "loss": 4.0671,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9546441442828476e-05,
+      "loss": 4.0593,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953807187412169e-05,
+      "loss": 4.0735,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952968592661117e-05,
+      "loss": 4.0739,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952129997910065e-05,
+      "loss": 4.0728,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9512914031590125e-05,
+      "loss": 4.0831,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950454446288334e-05,
+      "loss": 4.0647,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949615851537282e-05,
+      "loss": 4.0583,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9487772567862294e-05,
+      "loss": 4.0821,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9479386620351774e-05,
+      "loss": 4.0567,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947100067284126e-05,
+      "loss": 4.0627,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946261472533074e-05,
+      "loss": 4.0666,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945424515662395e-05,
+      "loss": 4.077,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944585920911343e-05,
+      "loss": 4.0556,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943747326160291e-05,
+      "loss": 4.0616,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942908731409239e-05,
+      "loss": 4.0544,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942070136658187e-05,
+      "loss": 4.0653,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941231541907135e-05,
+      "loss": 4.0762,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940394585036456e-05,
+      "loss": 4.0654,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939555990285404e-05,
+      "loss": 4.0664,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938717395534352e-05,
+      "loss": 4.0648,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9378788007833e-05,
+      "loss": 4.0869,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9370418439126214e-05,
+      "loss": 4.0563,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9362032491615694e-05,
+      "loss": 4.0685,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9353662922908903e-05,
+      "loss": 4.0691,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.934527697539838e-05,
+      "loss": 4.0607,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933689102788786e-05,
+      "loss": 4.0716,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932850508037734e-05,
+      "loss": 4.0676,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932011913286682e-05,
+      "loss": 4.0642,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93117331853563e-05,
+      "loss": 4.0688,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930334723784578e-05,
+      "loss": 4.0606,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929496129033526e-05,
+      "loss": 4.0632,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928657534282474e-05,
+      "loss": 4.065,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927820577411795e-05,
+      "loss": 4.0712,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926981982660743e-05,
+      "loss": 4.058,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926143387909691e-05,
+      "loss": 4.0604,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925304793158639e-05,
+      "loss": 4.057,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924466198407588e-05,
+      "loss": 4.072,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923627603656536e-05,
+      "loss": 4.0652,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922789008905484e-05,
+      "loss": 4.0515,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921952052034805e-05,
+      "loss": 4.0681,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921113457283753e-05,
+      "loss": 4.0725,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920274862532701e-05,
+      "loss": 4.0647,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919436267781648e-05,
+      "loss": 4.0471,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918597673030596e-05,
+      "loss": 4.0603,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917760716159918e-05,
+      "loss": 4.0561,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916922121408866e-05,
+      "loss": 4.0735,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916083526657813e-05,
+      "loss": 4.066,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9152449319067617e-05,
+      "loss": 4.0635,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9144063371557096e-05,
+      "loss": 4.0585,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913571018165404e-05,
+      "loss": 4.0569,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912732423414352e-05,
+      "loss": 4.0534,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9118938286633e-05,
+      "loss": 4.0572,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911055233912248e-05,
+      "loss": 4.0519,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9102166391611955e-05,
+      "loss": 4.0691,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9093780444101435e-05,
+      "loss": 4.0471,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9085394496590914e-05,
+      "loss": 4.0505,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9077008549080394e-05,
+      "loss": 4.0563,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9068622601569874e-05,
+      "loss": 4.0558,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9060253032863083e-05,
+      "loss": 4.0569,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905186708535257e-05,
+      "loss": 4.0512,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904348113784205e-05,
+      "loss": 4.063,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903509519033153e-05,
+      "loss": 4.0703,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902670924282101e-05,
+      "loss": 4.0493,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901833967411422e-05,
+      "loss": 4.0624,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90099537266037e-05,
+      "loss": 4.0552,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900156777909318e-05,
+      "loss": 4.0616,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899318183158266e-05,
+      "loss": 4.0567,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898479588407214e-05,
+      "loss": 4.0543,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897640993656162e-05,
+      "loss": 4.0534,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89680239890511e-05,
+      "loss": 4.0587,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895965442034431e-05,
+      "loss": 4.062,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895126847283379e-05,
+      "loss": 4.0562,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.894288252532327e-05,
+      "loss": 4.0524,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8934496577812755e-05,
+      "loss": 4.057,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8926110630302235e-05,
+      "loss": 4.0419,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917724682791715e-05,
+      "loss": 4.0694,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.0498,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8900952787770674e-05,
+      "loss": 4.0648,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889259959786761e-05,
+      "loss": 4.0466,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888421365035709e-05,
+      "loss": 4.0673,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887582770284657e-05,
+      "loss": 4.0602,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886744175533605e-05,
+      "loss": 4.0554,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885905580782553e-05,
+      "loss": 4.0576,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885068623911874e-05,
+      "loss": 4.0573,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884230029160822e-05,
+      "loss": 4.0586,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88339143440977e-05,
+      "loss": 4.0715,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882552839658719e-05,
+      "loss": 4.0395,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881714244907667e-05,
+      "loss": 4.0665,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880875650156615e-05,
+      "loss": 4.047,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880037055405562e-05,
+      "loss": 4.0606,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879200098534884e-05,
+      "loss": 4.055,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878361503783832e-05,
+      "loss": 4.0508,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877522909032779e-05,
+      "loss": 4.0579,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876684314281727e-05,
+      "loss": 4.0614,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875845719530675e-05,
+      "loss": 4.0563,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.130916595458984,
+      "eval_runtime": 292.3274,
+      "eval_samples_per_second": 1305.355,
+      "eval_steps_per_second": 40.793,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875007124779623e-05,
+      "loss": 4.0508,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874168530028571e-05,
+      "loss": 4.0528,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873329935277519e-05,
+      "loss": 4.0598,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872491340526467e-05,
+      "loss": 4.0555,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871652745775416e-05,
+      "loss": 4.0599,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870814151024364e-05,
+      "loss": 4.0513,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869975556273312e-05,
+      "loss": 4.0501,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86913696152226e-05,
+      "loss": 4.0495,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868298366771208e-05,
+      "loss": 4.0536,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674597720201557e-05,
+      "loss": 4.0623,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866621177269103e-05,
+      "loss": 4.0509,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865782582518051e-05,
+      "loss": 4.055,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8649456256473726e-05,
+      "loss": 4.0568,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641070308963206e-05,
+      "loss": 4.0443,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863268436145268e-05,
+      "loss": 4.0501,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.862429841394216e-05,
+      "loss": 4.04,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861591246643164e-05,
+      "loss": 4.0468,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8607526518921125e-05,
+      "loss": 4.0425,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599140571410605e-05,
+      "loss": 4.0448,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590754623900085e-05,
+      "loss": 4.0665,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8582368676389565e-05,
+      "loss": 4.0593,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8573982728879045e-05,
+      "loss": 4.0522,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565613160172254e-05,
+      "loss": 4.0501,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8557227212661734e-05,
+      "loss": 4.0606,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8548841265151214e-05,
+      "loss": 4.0363,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8540455317640694e-05,
+      "loss": 4.0515,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85320857489339e-05,
+      "loss": 4.0522,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852369980142338e-05,
+      "loss": 4.0362,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851531385391286e-05,
+      "loss": 4.0464,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850692790640234e-05,
+      "loss": 4.0385,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849854195889182e-05,
+      "loss": 4.0488,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849015601138131e-05,
+      "loss": 4.0522,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848178644267452e-05,
+      "loss": 4.0476,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473400495164e-05,
+      "loss": 4.0549,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846501454765348e-05,
+      "loss": 4.0441,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845662860014296e-05,
+      "loss": 4.0516,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844824265263244e-05,
+      "loss": 4.0546,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843985670512192e-05,
+      "loss": 4.0456,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843147075761139e-05,
+      "loss": 4.0382,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842310118890461e-05,
+      "loss": 4.0492,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841473162019782e-05,
+      "loss": 4.0391,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84063456726873e-05,
+      "loss": 4.0573,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839795972517678e-05,
+      "loss": 4.0517,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838957377766626e-05,
+      "loss": 4.0458,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8381187830155743e-05,
+      "loss": 4.0539,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8372801882645223e-05,
+      "loss": 4.0448,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8364415935134697e-05,
+      "loss": 4.0524,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8356029987624177e-05,
+      "loss": 4.0489,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8347644040113656e-05,
+      "loss": 4.0157,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8339258092603136e-05,
+      "loss": 4.0599,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8330872145092616e-05,
+      "loss": 4.0458,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8322502576385826e-05,
+      "loss": 4.0558,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314116628875305e-05,
+      "loss": 4.0427,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8305730681364785e-05,
+      "loss": 4.0432,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8297344733854265e-05,
+      "loss": 4.0336,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828897516514748e-05,
+      "loss": 4.0487,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828058921763696e-05,
+      "loss": 4.0469,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827220327012644e-05,
+      "loss": 4.0463,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826381732261592e-05,
+      "loss": 4.063,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82554313751054e-05,
+      "loss": 4.0427,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824704542759488e-05,
+      "loss": 4.0281,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823865948008436e-05,
+      "loss": 4.0609,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823027353257384e-05,
+      "loss": 4.0336,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822190396386705e-05,
+      "loss": 4.033,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821351801635653e-05,
+      "loss": 4.0481,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820514844764974e-05,
+      "loss": 4.0479,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819676250013922e-05,
+      "loss": 4.0329,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81883765526287e-05,
+      "loss": 4.0372,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817999060511818e-05,
+      "loss": 4.0284,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171604657607666e-05,
+      "loss": 4.0391,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163235088900875e-05,
+      "loss": 4.0526,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154849141390355e-05,
+      "loss": 4.0426,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146463193879835e-05,
+      "loss": 4.0445,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8138077246369315e-05,
+      "loss": 4.0371,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129691298858795e-05,
+      "loss": 4.0603,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8121321730152004e-05,
+      "loss": 4.0353,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112935782641484e-05,
+      "loss": 4.0471,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8104549835130964e-05,
+      "loss": 4.0409,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8096163887620444e-05,
+      "loss": 4.0377,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808779431891365e-05,
+      "loss": 4.047,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807940837140313e-05,
+      "loss": 4.0468,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807102242389262e-05,
+      "loss": 4.037,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80626364763821e-05,
+      "loss": 4.0484,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805425052887158e-05,
+      "loss": 4.0375,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804586458136105e-05,
+      "loss": 4.0385,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803747863385053e-05,
+      "loss": 4.0411,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802909268634001e-05,
+      "loss": 4.0463,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802070673882949e-05,
+      "loss": 4.035,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801232079131897e-05,
+      "loss": 4.0399,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800395122261218e-05,
+      "loss": 4.0346,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799556527510166e-05,
+      "loss": 4.041,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798717932759114e-05,
+      "loss": 4.0443,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797879338008062e-05,
+      "loss": 4.0279,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797042381137384e-05,
+      "loss": 4.0449,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796203786386332e-05,
+      "loss": 4.0421,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79536519163528e-05,
+      "loss": 4.0456,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794526596884228e-05,
+      "loss": 4.0217,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793688002133176e-05,
+      "loss": 4.0347,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792849407382124e-05,
+      "loss": 4.0337,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920124505114446e-05,
+      "loss": 4.0496,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911738557603926e-05,
+      "loss": 4.0436,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903352610093406e-05,
+      "loss": 4.0408,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7894966662582886e-05,
+      "loss": 4.0368,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7886580715072366e-05,
+      "loss": 4.0325,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7878194767561846e-05,
+      "loss": 4.0277,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7869825198855055e-05,
+      "loss": 4.0352,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7861439251344535e-05,
+      "loss": 4.0297,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785305330383402e-05,
+      "loss": 4.0399,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78446673563235e-05,
+      "loss": 4.0272,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783629778761671e-05,
+      "loss": 4.0254,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782791184010619e-05,
+      "loss": 4.0313,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781952589259567e-05,
+      "loss": 4.0363,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781113994508515e-05,
+      "loss": 4.0333,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780275399757463e-05,
+      "loss": 4.0267,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779436805006411e-05,
+      "loss": 4.0433,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778598210255359e-05,
+      "loss": 4.0443,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7777596155043063e-05,
+      "loss": 4.0293,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7769210207532543e-05,
+      "loss": 4.0361,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776084063882576e-05,
+      "loss": 4.0355,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775245469131524e-05,
+      "loss": 4.0361,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774406874380472e-05,
+      "loss": 4.0359,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7735699175097935e-05,
+      "loss": 4.0291,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7727313227587415e-05,
+      "loss": 4.0335,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771892728007689e-05,
+      "loss": 4.0323,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771054133256637e-05,
+      "loss": 4.0371,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770215538505585e-05,
+      "loss": 4.0328,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693785816349064e-05,
+      "loss": 4.0248,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768541624764227e-05,
+      "loss": 4.0338,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767703030013175e-05,
+      "loss": 4.0184,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766864435262123e-05,
+      "loss": 4.0463,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766025840511071e-05,
+      "loss": 4.0279,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765187245760019e-05,
+      "loss": 4.0417,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764348651008967e-05,
+      "loss": 4.0253,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763510056257915e-05,
+      "loss": 4.0437,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762671461506863e-05,
+      "loss": 4.0398,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761832866755811e-05,
+      "loss": 4.0273,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760995909885132e-05,
+      "loss": 4.0345,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76015731513408e-05,
+      "loss": 4.0313,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759318720383028e-05,
+      "loss": 4.0409,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758480125631976e-05,
+      "loss": 4.0455,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757641530880924e-05,
+      "loss": 4.0186,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756802936129872e-05,
+      "loss": 4.0386,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75596434137882e-05,
+      "loss": 4.0289,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755127384508141e-05,
+      "loss": 4.0371,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75428878975709e-05,
+      "loss": 4.0305,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753450195006038e-05,
+      "loss": 4.032,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752611600254986e-05,
+      "loss": 4.0334,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7517746433843066e-05,
+      "loss": 4.038,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7509360486332546e-05,
+      "loss": 4.0367,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.115942001342773,
+      "eval_runtime": 291.7495,
+      "eval_samples_per_second": 1307.941,
+      "eval_steps_per_second": 40.874,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7500974538822026e-05,
+      "loss": 4.0276,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7492588591311506e-05,
+      "loss": 4.0292,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7484202643800986e-05,
+      "loss": 4.0373,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7475833075094195e-05,
+      "loss": 4.0342,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7467447127583675e-05,
+      "loss": 4.0385,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7459061180073155e-05,
+      "loss": 4.0295,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7450675232562635e-05,
+      "loss": 4.0281,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7442289285052115e-05,
+      "loss": 4.0279,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7433903337541595e-05,
+      "loss": 4.0284,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7425517390031075e-05,
+      "loss": 4.0435,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7417131442520555e-05,
+      "loss": 4.0269,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740876187381377e-05,
+      "loss": 4.0284,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740039230510698e-05,
+      "loss": 4.0369,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739202273640019e-05,
+      "loss": 4.0212,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738363678888967e-05,
+      "loss": 4.0312,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737525084137915e-05,
+      "loss": 4.0172,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736686489386863e-05,
+      "loss": 4.0224,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735847894635811e-05,
+      "loss": 4.0253,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735009299884759e-05,
+      "loss": 4.0194,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734170705133707e-05,
+      "loss": 4.0441,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7333337482630285e-05,
+      "loss": 4.0404,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7324951535119765e-05,
+      "loss": 4.0302,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7316565587609245e-05,
+      "loss": 4.0293,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7308179640098725e-05,
+      "loss": 4.0327,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72997936925882e-05,
+      "loss": 4.0198,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7291424123881414e-05,
+      "loss": 4.0285,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7283038176370894e-05,
+      "loss": 4.0286,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7274652228860374e-05,
+      "loss": 4.013,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726626628134985e-05,
+      "loss": 4.023,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725788033383933e-05,
+      "loss": 4.0184,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724949438632881e-05,
+      "loss": 4.0301,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7241108438818287e-05,
+      "loss": 4.0271,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7232722491307767e-05,
+      "loss": 4.0257,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722433654379725e-05,
+      "loss": 4.0314,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721595059628673e-05,
+      "loss": 4.0203,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720756464877621e-05,
+      "loss": 4.0331,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719919508006942e-05,
+      "loss": 4.0371,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71908091325589e-05,
+      "loss": 4.0185,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718242318504838e-05,
+      "loss": 4.022,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717405361634159e-05,
+      "loss": 4.024,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716566766883107e-05,
+      "loss": 4.018,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715728172132055e-05,
+      "loss": 4.0351,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714889577381003e-05,
+      "loss": 4.0277,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714050982629951e-05,
+      "loss": 4.026,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713212387878899e-05,
+      "loss": 4.0306,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712373793127847e-05,
+      "loss": 4.0258,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711535198376795e-05,
+      "loss": 4.0286,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710696603625744e-05,
+      "loss": 4.0298,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709858008874691e-05,
+      "loss": 3.9979,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709019414123639e-05,
+      "loss": 4.0365,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708180819372587e-05,
+      "loss": 4.0209,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707342224621535e-05,
+      "loss": 4.0339,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706505267750856e-05,
+      "loss": 4.0237,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7056683108801776e-05,
+      "loss": 4.0199,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7048297161291256e-05,
+      "loss": 4.016,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703991121378073e-05,
+      "loss": 4.0264,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703152526627021e-05,
+      "loss": 4.0272,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702313931875969e-05,
+      "loss": 4.0202,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7014753371249176e-05,
+      "loss": 4.0414,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7006383802542385e-05,
+      "loss": 4.0261,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6997997855031865e-05,
+      "loss": 4.0012,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6989611907521345e-05,
+      "loss": 4.0354,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6981225960010824e-05,
+      "loss": 4.0127,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6972840012500304e-05,
+      "loss": 4.014,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6964454064989784e-05,
+      "loss": 4.0275,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6956068117479264e-05,
+      "loss": 4.0263,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6947698548772473e-05,
+      "loss": 4.0116,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6939312601261953e-05,
+      "loss": 4.0157,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693092665375143e-05,
+      "loss": 4.0071,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692254070624091e-05,
+      "loss": 4.0172,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691415475873039e-05,
+      "loss": 4.0289,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690578519002361e-05,
+      "loss": 4.0231,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689739924251309e-05,
+      "loss": 4.0248,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688901329500257e-05,
+      "loss": 4.0152,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688062734749205e-05,
+      "loss": 4.038,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687224139998153e-05,
+      "loss": 4.0188,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686385545247101e-05,
+      "loss": 4.0262,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685548588376422e-05,
+      "loss": 4.0162,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68470999362537e-05,
+      "loss": 4.0215,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683871398874318e-05,
+      "loss": 4.0259,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683032804123266e-05,
+      "loss": 4.0253,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682194209372214e-05,
+      "loss": 4.0156,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681355614621162e-05,
+      "loss": 4.0296,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68051701987011e-05,
+      "loss": 4.0093,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679678425119058e-05,
+      "loss": 4.0201,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6788414682483794e-05,
+      "loss": 4.0211,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6780028734973274e-05,
+      "loss": 4.0276,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677164278746275e-05,
+      "loss": 4.0154,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676325683995223e-05,
+      "loss": 4.0169,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675487089244171e-05,
+      "loss": 4.0138,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674650132373492e-05,
+      "loss": 4.0199,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6738115376224396e-05,
+      "loss": 4.023,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6729729428713876e-05,
+      "loss": 4.0095,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6721343481203356e-05,
+      "loss": 4.0232,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6712973912496565e-05,
+      "loss": 4.0255,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6704587964986045e-05,
+      "loss": 4.028,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669620201747553e-05,
+      "loss": 3.9993,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668781606996501e-05,
+      "loss": 4.0164,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667943012245449e-05,
+      "loss": 4.0148,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667104417494397e-05,
+      "loss": 4.0262,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666267460623718e-05,
+      "loss": 4.0268,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665428865872666e-05,
+      "loss": 4.0158,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664590271121614e-05,
+      "loss": 4.0135,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663753314250935e-05,
+      "loss": 4.0162,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662914719499883e-05,
+      "loss": 4.0083,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662076124748831e-05,
+      "loss": 4.0144,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661237529997779e-05,
+      "loss": 4.0105,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660398935246727e-05,
+      "loss": 4.0209,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659560340495675e-05,
+      "loss": 4.0069,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658721745744623e-05,
+      "loss": 4.0076,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6578831509935716e-05,
+      "loss": 4.0079,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6570461941228925e-05,
+      "loss": 4.0135,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6562075993718405e-05,
+      "loss": 4.0159,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6553690046207885e-05,
+      "loss": 4.0065,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6545320477501094e-05,
+      "loss": 4.0181,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6536934529990574e-05,
+      "loss": 4.0189,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6528548582480054e-05,
+      "loss": 4.0119,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6520162634969534e-05,
+      "loss": 4.0187,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6511776687459014e-05,
+      "loss": 4.0136,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6503390739948494e-05,
+      "loss": 4.0171,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6495004792437974e-05,
+      "loss": 4.0125,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6486618844927454e-05,
+      "loss": 4.009,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647824927622067e-05,
+      "loss": 4.0165,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646986332871015e-05,
+      "loss": 4.0096,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646147738119963e-05,
+      "loss": 4.019,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645309143368911e-05,
+      "loss": 4.0152,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644472186498232e-05,
+      "loss": 4.0051,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64363359174718e-05,
+      "loss": 4.0095,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642794996996128e-05,
+      "loss": 4.0008,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641956402245075e-05,
+      "loss": 4.023,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641117807494023e-05,
+      "loss": 4.0064,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640279212742971e-05,
+      "loss": 4.0202,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639442255872293e-05,
+      "loss": 4.0088,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638603661121241e-05,
+      "loss": 4.0224,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637765066370189e-05,
+      "loss": 4.0208,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636926471619137e-05,
+      "loss": 4.0075,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636087876868085e-05,
+      "loss": 4.0176,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635249282117033e-05,
+      "loss": 4.0088,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6344123252463536e-05,
+      "loss": 4.0215,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6335737304953016e-05,
+      "loss": 4.0203,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6327351357442496e-05,
+      "loss": 4.0058,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6318965409931976e-05,
+      "loss": 4.0152,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6310579462421456e-05,
+      "loss": 4.0114,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6302193514910936e-05,
+      "loss": 4.0208,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6293807567400416e-05,
+      "loss": 4.0068,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6285437998693625e-05,
+      "loss": 4.0194,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6277052051183105e-05,
+      "loss": 4.0072,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626866610367259e-05,
+      "loss": 4.0199,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.62602965349658e-05,
+      "loss": 4.0134,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.10367488861084,
+      "eval_runtime": 291.8187,
+      "eval_samples_per_second": 1307.63,
+      "eval_steps_per_second": 40.864,
+      "step": 839520
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.625191058745528e-05,
+      "loss": 4.0103,
+      "step": 839680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.624352463994476e-05,
+      "loss": 4.0091,
+      "step": 840192
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.623513869243424e-05,
+      "loss": 4.0141,
+      "step": 840704
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.622675274492372e-05,
+      "loss": 4.0179,
+      "step": 841216
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.62183667974132e-05,
+      "loss": 4.0189,
+      "step": 841728
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.620998084990268e-05,
+      "loss": 4.0087,
+      "step": 842240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.620159490239216e-05,
+      "loss": 4.0081,
+      "step": 842752
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.619320895488164e-05,
+      "loss": 4.0075,
+      "step": 843264
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.618483938617485e-05,
+      "loss": 4.0115,
+      "step": 843776
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.617645343866433e-05,
+      "loss": 4.0188,
+      "step": 844288
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.616806749115381e-05,
+      "loss": 4.0065,
+      "step": 844800
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6159697922447025e-05,
+      "loss": 4.0107,
+      "step": 845312
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6151328353740234e-05,
+      "loss": 4.0215,
+      "step": 845824
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6142942406229714e-05,
+      "loss": 4.0028,
+      "step": 846336
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6134556458719194e-05,
+      "loss": 4.0054,
+      "step": 846848
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6126170511208674e-05,
+      "loss": 3.9985,
+      "step": 847360
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6117784563698154e-05,
+      "loss": 4.0012,
+      "step": 847872
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6109398616187634e-05,
+      "loss": 4.0085,
+      "step": 848384
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6101012668677114e-05,
+      "loss": 4.0064,
+      "step": 848896
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.609262672116659e-05,
+      "loss": 4.0183,
+      "step": 849408
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.608424077365607e-05,
+      "loss": 4.0214,
+      "step": 849920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.607585482614555e-05,
+      "loss": 4.0102,
+      "step": 850432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.606746887863503e-05,
+      "loss": 4.0103,
+      "step": 850944
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.605908293112451e-05,
+      "loss": 4.0141,
+      "step": 851456
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6050696983613994e-05,
+      "loss": 3.9971,
+      "step": 851968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6042311036103474e-05,
+      "loss": 4.0142,
+      "step": 852480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6033925088592954e-05,
+      "loss": 4.005,
+      "step": 852992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.6025539141082434e-05,
+      "loss": 3.9937,
+      "step": 853504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.601716957237564e-05,
+      "loss": 4.0043,
+      "step": 854016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.600878362486512e-05,
+      "loss": 4.002,
+      "step": 854528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.600041405615833e-05,
+      "loss": 4.0053,
+      "step": 855040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.599202810864781e-05,
+      "loss": 4.0073,
+      "step": 855552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.598364216113729e-05,
+      "loss": 4.012,
+      "step": 856064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.597525621362677e-05,
+      "loss": 4.0098,
+      "step": 856576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.596688664491998e-05,
+      "loss": 3.9992,
+      "step": 857088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.595850069740946e-05,
+      "loss": 4.0136,
+      "step": 857600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.595011474989895e-05,
+      "loss": 4.0176,
+      "step": 858112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.594172880238843e-05,
+      "loss": 4.0011,
+      "step": 858624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.593334285487791e-05,
+      "loss": 4.0033,
+      "step": 859136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.592495690736739e-05,
+      "loss": 4.0016,
+      "step": 859648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.591657095985687e-05,
+      "loss": 3.9962,
+      "step": 860160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.590818501234635e-05,
+      "loss": 4.0135,
+      "step": 860672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5899815443639556e-05,
+      "loss": 4.0095,
+      "step": 861184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5891429496129036e-05,
+      "loss": 4.0137,
+      "step": 861696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5883043548618516e-05,
+      "loss": 4.0061,
+      "step": 862208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5874657601107996e-05,
+      "loss": 4.0074,
+      "step": 862720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5866271653597476e-05,
+      "loss": 4.0094,
+      "step": 863232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.585788570608695e-05,
+      "loss": 4.0077,
+      "step": 863744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.584949975857643e-05,
+      "loss": 3.9817,
+      "step": 864256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5841113811065916e-05,
+      "loss": 4.0169,
+      "step": 864768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.583274424235913e-05,
+      "loss": 4.0013,
+      "step": 865280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.582437467365234e-05,
+      "loss": 4.0171,
+      "step": 865792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.581598872614182e-05,
+      "loss": 4.0022,
+      "step": 866304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.58076027786313e-05,
+      "loss": 4.0019,
+      "step": 866816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5799216831120774e-05,
+      "loss": 3.9966,
+      "step": 867328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.579084726241399e-05,
+      "loss": 4.0041,
+      "step": 867840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.578246131490347e-05,
+      "loss": 4.0132,
+      "step": 868352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.577407536739295e-05,
+      "loss": 4.0006,
+      "step": 868864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.576568941988242e-05,
+      "loss": 4.02,
+      "step": 869376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.57573034723719e-05,
+      "loss": 4.0059,
+      "step": 869888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.574893390366512e-05,
+      "loss": 3.9842,
+      "step": 870400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.57405479561546e-05,
+      "loss": 4.0172,
+      "step": 870912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.573216200864408e-05,
+      "loss": 3.9965,
+      "step": 871424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.572377606113356e-05,
+      "loss": 3.9966,
+      "step": 871936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.571539011362304e-05,
+      "loss": 4.006,
+      "step": 872448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.570700416611252e-05,
+      "loss": 4.0084,
+      "step": 872960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5698618218602e-05,
+      "loss": 3.9916,
+      "step": 873472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.569023227109148e-05,
+      "loss": 4.0009,
+      "step": 873984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.568186270238469e-05,
+      "loss": 3.9874,
+      "step": 874496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.567347675487417e-05,
+      "loss": 4.0046,
+      "step": 875008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.566509080736365e-05,
+      "loss": 4.0068,
+      "step": 875520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.565670485985313e-05,
+      "loss": 4.0014,
+      "step": 876032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.564833529114634e-05,
+      "loss": 4.0083,
+      "step": 876544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5639949343635823e-05,
+      "loss": 3.9967,
+      "step": 877056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5631563396125303e-05,
+      "loss": 4.0164,
+      "step": 877568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.562317744861478e-05,
+      "loss": 4.0043,
+      "step": 878080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.561479150110426e-05,
+      "loss": 4.0093,
+      "step": 878592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.560642193239747e-05,
+      "loss": 3.9985,
+      "step": 879104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.559803598488695e-05,
+      "loss": 4.0023,
+      "step": 879616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.558965003737643e-05,
+      "loss": 4.0059,
+      "step": 880128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.558126408986591e-05,
+      "loss": 4.0058,
+      "step": 880640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.557289452115912e-05,
+      "loss": 4.0003,
+      "step": 881152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.55645085736486e-05,
+      "loss": 4.009,
+      "step": 881664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.555612262613808e-05,
+      "loss": 3.9917,
+      "step": 882176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.554775305743129e-05,
+      "loss": 3.9984,
+      "step": 882688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.553936710992077e-05,
+      "loss": 3.9995,
+      "step": 883200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.553098116241026e-05,
+      "loss": 4.0117,
+      "step": 883712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.552259521489974e-05,
+      "loss": 3.994,
+      "step": 884224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.551420926738922e-05,
+      "loss": 3.9996,
+      "step": 884736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.5505839698682426e-05,
+      "loss": 3.9974,
+      "step": 885248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5497453751171906e-05,
+      "loss": 4.0005,
+      "step": 885760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5489067803661386e-05,
+      "loss": 4.0037,
+      "step": 886272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5480681856150866e-05,
+      "loss": 3.9912,
+      "step": 886784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5472295908640346e-05,
+      "loss": 4.0033,
+      "step": 887296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5463926339933555e-05,
+      "loss": 4.0087,
+      "step": 887808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5455540392423035e-05,
+      "loss": 4.0076,
+      "step": 888320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5447154444912515e-05,
+      "loss": 3.9838,
+      "step": 888832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5438768497401995e-05,
+      "loss": 3.9952,
+      "step": 889344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5430382549891475e-05,
+      "loss": 3.9936,
+      "step": 889856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.542201298118469e-05,
+      "loss": 4.0079,
+      "step": 890368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.541362703367417e-05,
+      "loss": 4.0098,
+      "step": 890880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.540524108616365e-05,
+      "loss": 3.997,
+      "step": 891392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.539685513865313e-05,
+      "loss": 3.9945,
+      "step": 891904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.538846919114261e-05,
+      "loss": 3.9986,
+      "step": 892416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5380083243632084e-05,
+      "loss": 3.9893,
+      "step": 892928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5371697296121564e-05,
+      "loss": 3.9955,
+      "step": 893440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.536332772741478e-05,
+      "loss": 3.9944,
+      "step": 893952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.535494177990426e-05,
+      "loss": 4.0003,
+      "step": 894464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.534655583239373e-05,
+      "loss": 3.9923,
+      "step": 894976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.533816988488321e-05,
+      "loss": 3.9863,
+      "step": 895488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.532980031617643e-05,
+      "loss": 3.9898,
+      "step": 896000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.532141436866591e-05,
+      "loss": 3.9933,
+      "step": 896512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.531302842115539e-05,
+      "loss": 3.9994,
+      "step": 897024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.530464247364487e-05,
+      "loss": 3.9888,
+      "step": 897536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.529625652613435e-05,
+      "loss": 4.0013,
+      "step": 898048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.528788695742756e-05,
+      "loss": 3.9994,
+      "step": 898560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.527950100991704e-05,
+      "loss": 3.9967,
+      "step": 899072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.527111506240652e-05,
+      "loss": 4.0015,
+      "step": 899584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5262729114896e-05,
+      "loss": 3.9933,
+      "step": 900096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.525434316738548e-05,
+      "loss": 4.0022,
+      "step": 900608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.524595721987496e-05,
+      "loss": 3.994,
+      "step": 901120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.523757127236444e-05,
+      "loss": 3.9922,
+      "step": 901632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.522918532485392e-05,
+      "loss": 3.999,
+      "step": 902144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.52207993773434e-05,
+      "loss": 3.9945,
+      "step": 902656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.521242980863661e-05,
+      "loss": 4.0008,
+      "step": 903168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.520404386112609e-05,
+      "loss": 3.9956,
+      "step": 903680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.519565791361557e-05,
+      "loss": 3.9899,
+      "step": 904192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.518728834490878e-05,
+      "loss": 3.9901,
+      "step": 904704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.517890239739826e-05,
+      "loss": 3.9851,
+      "step": 905216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.517051644988774e-05,
+      "loss": 4.0031,
+      "step": 905728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.516213050237722e-05,
+      "loss": 3.9944,
+      "step": 906240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.51537445548667e-05,
+      "loss": 4.0039,
+      "step": 906752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.514535860735618e-05,
+      "loss": 3.9893,
+      "step": 907264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.513697265984566e-05,
+      "loss": 4.0035,
+      "step": 907776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.512858671233514e-05,
+      "loss": 4.0018,
+      "step": 908288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.512021714362835e-05,
+      "loss": 3.9929,
+      "step": 908800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.511183119611783e-05,
+      "loss": 3.9957,
+      "step": 909312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.510344524860731e-05,
+      "loss": 3.9927,
+      "step": 909824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.50950593010968e-05,
+      "loss": 4.004,
+      "step": 910336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.508667335358627e-05,
+      "loss": 4.0031,
+      "step": 910848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.507828740607575e-05,
+      "loss": 3.9901,
+      "step": 911360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.506990145856523e-05,
+      "loss": 3.9905,
+      "step": 911872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.506151551105471e-05,
+      "loss": 3.9978,
+      "step": 912384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.505314594234792e-05,
+      "loss": 4.0015,
+      "step": 912896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5044776373641135e-05,
+      "loss": 3.9882,
+      "step": 913408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5036406804934344e-05,
+      "loss": 3.9998,
+      "step": 913920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5028020857423824e-05,
+      "loss": 3.9903,
+      "step": 914432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5019634909913304e-05,
+      "loss": 4.0037,
+      "step": 914944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.5011248962402784e-05,
+      "loss": 3.9977,
+      "step": 915456
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.092974662780762,
+      "eval_runtime": 292.9858,
+      "eval_samples_per_second": 1302.422,
+      "eval_steps_per_second": 40.702,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002863014892264e-05,
+      "loss": 3.9998,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4994477067381744e-05,
+      "loss": 3.9899,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4986091119871224e-05,
+      "loss": 3.9964,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4977705172360704e-05,
+      "loss": 4.0015,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4969319224850184e-05,
+      "loss": 4.0009,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4960933277339664e-05,
+      "loss": 3.9951,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952547329829144e-05,
+      "loss": 3.9913,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4944161382318624e-05,
+      "loss": 3.9934,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4935775434808104e-05,
+      "loss": 3.9903,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4927389487297584e-05,
+      "loss": 3.9977,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4919003539787064e-05,
+      "loss": 3.9946,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4910617592276544e-05,
+      "loss": 3.9923,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490224802356975e-05,
+      "loss": 4.0051,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489386207605923e-05,
+      "loss": 3.9841,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488547612854872e-05,
+      "loss": 3.9881,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48770901810382e-05,
+      "loss": 3.9833,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486870423352768e-05,
+      "loss": 3.9826,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486031828601716e-05,
+      "loss": 3.9915,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485193233850663e-05,
+      "loss": 3.9883,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484354639099611e-05,
+      "loss": 4.0002,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483516044348559e-05,
+      "loss": 4.0057,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482679087477881e-05,
+      "loss": 3.9946,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481840492726828e-05,
+      "loss": 3.9951,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481001897975776e-05,
+      "loss": 3.9959,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480163303224724e-05,
+      "loss": 3.9797,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479326346354046e-05,
+      "loss": 4.0007,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478487751602994e-05,
+      "loss": 3.9868,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477649156851942e-05,
+      "loss": 3.9791,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47681056210089e-05,
+      "loss": 3.9901,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475971967349838e-05,
+      "loss": 3.9819,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751350104791586e-05,
+      "loss": 3.9888,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4742964157281066e-05,
+      "loss": 3.9881,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734578209770546e-05,
+      "loss": 4.0002,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726192262260026e-05,
+      "loss": 3.9892,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717806314749506e-05,
+      "loss": 3.9861,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709420367238986e-05,
+      "loss": 3.9938,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4701034419728466e-05,
+      "loss": 3.9999,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4692648472217946e-05,
+      "loss": 3.9856,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4684278903511155e-05,
+      "loss": 3.9849,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467589295600064e-05,
+      "loss": 3.9853,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466750700849012e-05,
+      "loss": 3.9833,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46591210609796e-05,
+      "loss": 3.9919,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465075149227281e-05,
+      "loss": 3.9917,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464238192356602e-05,
+      "loss": 3.9998,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46339959760555e-05,
+      "loss": 3.9908,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462562640734871e-05,
+      "loss": 3.9926,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461724045983819e-05,
+      "loss": 3.9932,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460885451232767e-05,
+      "loss": 3.9919,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460046856481715e-05,
+      "loss": 3.9644,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459208261730663e-05,
+      "loss": 3.9965,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458369666979611e-05,
+      "loss": 3.9883,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575310722285595e-05,
+      "loss": 3.9997,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566924774775075e-05,
+      "loss": 3.9829,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558555206068285e-05,
+      "loss": 3.9901,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550169258557764e-05,
+      "loss": 3.9775,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.9848,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533397363536724e-05,
+      "loss": 4.0012,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525011416026204e-05,
+      "loss": 3.9855,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516625468515684e-05,
+      "loss": 4.0033,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508239521005164e-05,
+      "loss": 3.9897,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4499853573494644e-05,
+      "loss": 3.9665,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449150038359158e-05,
+      "loss": 4.006,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448311443608106e-05,
+      "loss": 3.9773,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447472848857054e-05,
+      "loss": 3.9788,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 3.9933,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 3.9906,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444957064603899e-05,
+      "loss": 3.9783,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444118469852847e-05,
+      "loss": 3.9811,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443281512982168e-05,
+      "loss": 3.9741,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442442918231116e-05,
+      "loss": 3.9809,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441604323480064e-05,
+      "loss": 3.9943,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440765728729012e-05,
+      "loss": 3.9878,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439927133977959e-05,
+      "loss": 3.9893,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439088539226907e-05,
+      "loss": 3.9808,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438249944475855e-05,
+      "loss": 3.9993,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437411349724803e-05,
+      "loss": 3.9883,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436572754973751e-05,
+      "loss": 3.9917,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435735798103073e-05,
+      "loss": 3.9815,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434897203352021e-05,
+      "loss": 3.9866,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434058608600969e-05,
+      "loss": 3.9934,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433220013849917e-05,
+      "loss": 3.9871,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432381419098865e-05,
+      "loss": 3.9877,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315444622281856e-05,
+      "loss": 3.9939,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307058674771336e-05,
+      "loss": 3.9792,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298672727260816e-05,
+      "loss": 3.9816,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4290286779750296e-05,
+      "loss": 3.9848,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281917211043505e-05,
+      "loss": 3.9935,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4273531263532985e-05,
+      "loss": 3.9803,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265145316022465e-05,
+      "loss": 3.9871,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425675936851195e-05,
+      "loss": 3.98,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424837342100143e-05,
+      "loss": 3.9829,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423998747349091e-05,
+      "loss": 3.9894,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423160152598039e-05,
+      "loss": 3.9779,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422321557846987e-05,
+      "loss": 3.9852,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421484600976308e-05,
+      "loss": 3.9923,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420646006225256e-05,
+      "loss": 3.9967,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419807411474204e-05,
+      "loss": 3.9614,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418968816723152e-05,
+      "loss": 3.9812,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181302219721e-05,
+      "loss": 3.9764,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417293265101421e-05,
+      "loss": 3.9929,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416456308230742e-05,
+      "loss": 3.9951,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156177134796905e-05,
+      "loss": 3.9796,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147791187286385e-05,
+      "loss": 3.9788,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139405239775865e-05,
+      "loss": 3.9837,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131019292265345e-05,
+      "loss": 3.976,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122633344754825e-05,
+      "loss": 3.9794,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114263776048034e-05,
+      "loss": 3.9766,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105877828537514e-05,
+      "loss": 3.9838,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097491881026994e-05,
+      "loss": 3.9752,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4089105933516474e-05,
+      "loss": 3.9732,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408071998600595e-05,
+      "loss": 3.9767,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407233403849543e-05,
+      "loss": 3.9723,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406394809098491e-05,
+      "loss": 3.9858,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405556214347439e-05,
+      "loss": 3.9704,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047176195963874e-05,
+      "loss": 3.9858,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403880662725708e-05,
+      "loss": 3.9831,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403042067974656e-05,
+      "loss": 3.983,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9821,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401366516352925e-05,
+      "loss": 3.9788,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400527921601873e-05,
+      "loss": 3.9872,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399689326850821e-05,
+      "loss": 3.9772,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398850732099769e-05,
+      "loss": 3.9755,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39801377522909e-05,
+      "loss": 3.9826,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397175180478038e-05,
+      "loss": 3.9783,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396336585726986e-05,
+      "loss": 3.9847,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395497990975934e-05,
+      "loss": 3.9785,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394659396224882e-05,
+      "loss": 3.9749,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393820801473831e-05,
+      "loss": 3.9769,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929838446031516e-05,
+      "loss": 3.9726,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921452498520996e-05,
+      "loss": 3.981,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913066551010476e-05,
+      "loss": 3.9795,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904680603499956e-05,
+      "loss": 3.9903,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896294655989436e-05,
+      "loss": 3.975,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3887908708478916e-05,
+      "loss": 3.9821,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879522760968396e-05,
+      "loss": 3.9858,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871136813457876e-05,
+      "loss": 3.9781,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862767244751085e-05,
+      "loss": 3.9767,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854381297240565e-05,
+      "loss": 3.9792,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3845995349730045e-05,
+      "loss": 3.9881,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3837609402219525e-05,
+      "loss": 3.9875,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382923983351274e-05,
+      "loss": 3.9781,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382085388600222e-05,
+      "loss": 3.9751,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381248431729543e-05,
+      "loss": 3.983,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380409836978491e-05,
+      "loss": 3.9842,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379572880107812e-05,
+      "loss": 3.9702,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37873428535676e-05,
+      "loss": 3.9916,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377895690605708e-05,
+      "loss": 3.9717,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 3.9874,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376218501103604e-05,
+      "loss": 3.9827,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.083831310272217,
+      "eval_runtime": 300.3073,
+      "eval_samples_per_second": 1270.669,
+      "eval_steps_per_second": 39.709,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375379906352552e-05,
+      "loss": 3.981,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3745413116015e-05,
+      "loss": 3.9741,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373702716850448e-05,
+      "loss": 3.9779,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728657599797694e-05,
+      "loss": 3.9901,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720271652287174e-05,
+      "loss": 3.9797,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3711885704776654e-05,
+      "loss": 3.9823,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703499757266134e-05,
+      "loss": 3.9751,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695113809755614e-05,
+      "loss": 3.9724,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368672786224509e-05,
+      "loss": 3.9766,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367834191473457e-05,
+      "loss": 3.9859,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366995596722405e-05,
+      "loss": 3.9751,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3661586398517256e-05,
+      "loss": 3.9849,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365321682981047e-05,
+      "loss": 3.9886,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364483088229995e-05,
+      "loss": 3.968,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363644493478943e-05,
+      "loss": 3.9774,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362805898727891e-05,
+      "loss": 3.9627,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361967303976839e-05,
+      "loss": 3.9648,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361128709225787e-05,
+      "loss": 3.9798,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360290114474735e-05,
+      "loss": 3.9698,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359451519723683e-05,
+      "loss": 3.9831,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358612924972631e-05,
+      "loss": 3.9913,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357774330221579e-05,
+      "loss": 3.9769,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356935735470527e-05,
+      "loss": 3.9826,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356097140719475e-05,
+      "loss": 3.9809,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355260183848796e-05,
+      "loss": 3.9664,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354421589097744e-05,
+      "loss": 3.9814,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353582994346692e-05,
+      "loss": 3.973,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35274439959564e-05,
+      "loss": 3.9682,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351905804844588e-05,
+      "loss": 3.9771,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35106884797391e-05,
+      "loss": 3.9641,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350230253222858e-05,
+      "loss": 3.9733,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3493932963521786e-05,
+      "loss": 3.9765,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485547016011266e-05,
+      "loss": 3.9825,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477161068500746e-05,
+      "loss": 3.973,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468775120990226e-05,
+      "loss": 3.9715,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460389173479706e-05,
+      "loss": 3.9741,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3452003225969185e-05,
+      "loss": 3.9873,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443617278458665e-05,
+      "loss": 3.9725,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435231330948145e-05,
+      "loss": 3.9696,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342687814104509e-05,
+      "loss": 3.9697,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341849219353457e-05,
+      "loss": 3.9738,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341010624602405e-05,
+      "loss": 3.974,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340172029851353e-05,
+      "loss": 3.9787,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339333435100301e-05,
+      "loss": 3.9829,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338494840349249e-05,
+      "loss": 3.9766,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337656245598197e-05,
+      "loss": 3.9762,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336817650847144e-05,
+      "loss": 3.9749,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335980693976466e-05,
+      "loss": 3.9757,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335143737105787e-05,
+      "loss": 3.9567,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334305142354735e-05,
+      "loss": 3.9793,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333466547603683e-05,
+      "loss": 3.9738,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332627952852631e-05,
+      "loss": 3.9837,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331789358101579e-05,
+      "loss": 3.9697,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309524012309004e-05,
+      "loss": 3.9727,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301138064798484e-05,
+      "loss": 3.9697,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292752117287964e-05,
+      "loss": 3.9645,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284366169777444e-05,
+      "loss": 3.9861,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327598022226692e-05,
+      "loss": 3.9681,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32675942747564e-05,
+      "loss": 3.9942,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325920832724588e-05,
+      "loss": 3.9713,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325082237973536e-05,
+      "loss": 3.9559,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32424691898323e-05,
+      "loss": 3.991,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323408324232178e-05,
+      "loss": 3.965,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322569729481126e-05,
+      "loss": 3.9613,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321731134730074e-05,
+      "loss": 3.973,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320892539979022e-05,
+      "loss": 3.9814,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32005394522797e-05,
+      "loss": 3.9612,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319216988357292e-05,
+      "loss": 3.97,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318378393606239e-05,
+      "loss": 3.9593,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317539798855187e-05,
+      "loss": 3.9625,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316701204104135e-05,
+      "loss": 3.9842,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3158642472334567e-05,
+      "loss": 3.9745,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315025652482404e-05,
+      "loss": 3.9724,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314187057731352e-05,
+      "loss": 3.9657,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133484629803e-05,
+      "loss": 3.9843,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312509868229248e-05,
+      "loss": 3.9734,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311671273478196e-05,
+      "loss": 3.9796,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108343166075175e-05,
+      "loss": 3.9664,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3099957218564655e-05,
+      "loss": 3.9706,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3091571271054135e-05,
+      "loss": 3.9794,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3083185323543615e-05,
+      "loss": 3.971,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3074799376033095e-05,
+      "loss": 3.972,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3066413428522575e-05,
+      "loss": 3.9799,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058027481012055e-05,
+      "loss": 3.97,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049641533501535e-05,
+      "loss": 3.9653,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3041271964794744e-05,
+      "loss": 3.9698,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3032886017284224e-05,
+      "loss": 3.9775,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302451644857743e-05,
+      "loss": 3.9683,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301613050106691e-05,
+      "loss": 3.9715,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30077445535564e-05,
+      "loss": 3.9638,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299935860604588e-05,
+      "loss": 3.9681,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299098903733909e-05,
+      "loss": 3.9726,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29826194686323e-05,
+      "loss": 3.9655,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297423352112178e-05,
+      "loss": 3.9702,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296584757361126e-05,
+      "loss": 3.9762,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295746162610074e-05,
+      "loss": 3.9833,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294907567859022e-05,
+      "loss": 3.9496,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29406897310797e-05,
+      "loss": 3.9663,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293230378356918e-05,
+      "loss": 3.9654,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292391783605866e-05,
+      "loss": 3.9773,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291553188854814e-05,
+      "loss": 3.9807,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290714594103762e-05,
+      "loss": 3.964,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28987599935271e-05,
+      "loss": 3.9642,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890374046016584e-05,
+      "loss": 3.9698,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2882004477309793e-05,
+      "loss": 3.9665,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2873618529799273e-05,
+      "loss": 3.9643,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286524896109248e-05,
+      "loss": 3.9624,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285686301358196e-05,
+      "loss": 3.9715,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284847706607144e-05,
+      "loss": 3.9616,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284009111856092e-05,
+      "loss": 3.9606,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28317051710504e-05,
+      "loss": 3.9644,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823319223539876e-05,
+      "loss": 3.9561,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2814933276029355e-05,
+      "loss": 3.9713,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2806547328518835e-05,
+      "loss": 3.963,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279816138100832e-05,
+      "loss": 3.9656,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278979181230153e-05,
+      "loss": 3.9713,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278140586479101e-05,
+      "loss": 3.9726,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277303629608423e-05,
+      "loss": 3.9671,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27646503485737e-05,
+      "loss": 3.9631,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275626440106318e-05,
+      "loss": 3.9762,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274787845355266e-05,
+      "loss": 3.9627,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273949250604214e-05,
+      "loss": 3.9593,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273112293733535e-05,
+      "loss": 3.972,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272273698982483e-05,
+      "loss": 3.9662,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271435104231431e-05,
+      "loss": 3.9673,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270596509480379e-05,
+      "loss": 3.9616,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269757914729327e-05,
+      "loss": 3.9669,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2689193199782756e-05,
+      "loss": 3.9633,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2680807252272236e-05,
+      "loss": 3.9593,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2672421304761716e-05,
+      "loss": 3.9584,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2664051736054925e-05,
+      "loss": 3.9728,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2655665788544405e-05,
+      "loss": 3.9763,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2647296219837614e-05,
+      "loss": 3.9581,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2638910272327094e-05,
+      "loss": 3.965,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2630524324816574e-05,
+      "loss": 3.9772,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2622138377306054e-05,
+      "loss": 3.9665,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2613752429795534e-05,
+      "loss": 3.9623,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2605366482285014e-05,
+      "loss": 3.9638,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259699691357822e-05,
+      "loss": 3.9781,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258862734487144e-05,
+      "loss": 3.9745,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258024139736092e-05,
+      "loss": 3.9635,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25718554498504e-05,
+      "loss": 3.9582,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256346950233988e-05,
+      "loss": 3.9693,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255508355482936e-05,
+      "loss": 3.9703,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254669760731884e-05,
+      "loss": 3.9578,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253831165980832e-05,
+      "loss": 3.978,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25299257122978e-05,
+      "loss": 3.9596,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252153976478728e-05,
+      "loss": 3.9745,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251317019608049e-05,
+      "loss": 3.9706,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.076541900634766,
+      "eval_runtime": 296.6287,
+      "eval_samples_per_second": 1286.426,
+      "eval_steps_per_second": 40.202,
+      "step": 1068480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.250478424856997e-05,
+      "loss": 3.9676,
+      "step": 1068544
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.249639830105945e-05,
+      "loss": 3.9575,
+      "step": 1069056
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.248801235354893e-05,
+      "loss": 3.9659,
+      "step": 1069568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.247962640603841e-05,
+      "loss": 3.9769,
+      "step": 1070080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.247125683733162e-05,
+      "loss": 3.9665,
+      "step": 1070592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.24628708898211e-05,
+      "loss": 3.9683,
+      "step": 1071104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.245448494231058e-05,
+      "loss": 3.9649,
+      "step": 1071616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.244609899480006e-05,
+      "loss": 3.9622,
+      "step": 1072128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2437713047289536e-05,
+      "loss": 3.9615,
+      "step": 1072640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2429327099779016e-05,
+      "loss": 3.9725,
+      "step": 1073152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.242095753107223e-05,
+      "loss": 3.962,
+      "step": 1073664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.241257158356171e-05,
+      "loss": 3.9722,
+      "step": 1074176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.240420201485492e-05,
+      "loss": 3.9736,
+      "step": 1074688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.23958160673444e-05,
+      "loss": 3.9531,
+      "step": 1075200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.238743011983388e-05,
+      "loss": 3.9633,
+      "step": 1075712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.237904417232336e-05,
+      "loss": 3.9513,
+      "step": 1076224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.237065822481284e-05,
+      "loss": 3.9549,
+      "step": 1076736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.236227227730232e-05,
+      "loss": 3.9644,
+      "step": 1077248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.23538863297918e-05,
+      "loss": 3.957,
+      "step": 1077760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.234550038228128e-05,
+      "loss": 3.9681,
+      "step": 1078272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.233711443477076e-05,
+      "loss": 3.9813,
+      "step": 1078784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.232872848726024e-05,
+      "loss": 3.9621,
+      "step": 1079296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.232034253974972e-05,
+      "loss": 3.9693,
+      "step": 1079808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.23119565922392e-05,
+      "loss": 3.967,
+      "step": 1080320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.230357064472868e-05,
+      "loss": 3.9531,
+      "step": 1080832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.229520107602189e-05,
+      "loss": 3.9716,
+      "step": 1081344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.228681512851137e-05,
+      "loss": 3.9565,
+      "step": 1081856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.227842918100085e-05,
+      "loss": 3.9592,
+      "step": 1082368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.227004323349033e-05,
+      "loss": 3.9595,
+      "step": 1082880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2261673664783545e-05,
+      "loss": 3.9505,
+      "step": 1083392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2253287717273025e-05,
+      "loss": 3.9587,
+      "step": 1083904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2244918148566234e-05,
+      "loss": 3.9634,
+      "step": 1084416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2236532201055714e-05,
+      "loss": 3.9695,
+      "step": 1084928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2228146253545194e-05,
+      "loss": 3.962,
+      "step": 1085440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2219760306034674e-05,
+      "loss": 3.9562,
+      "step": 1085952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2211374358524154e-05,
+      "loss": 3.9591,
+      "step": 1086464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2202988411013634e-05,
+      "loss": 3.9785,
+      "step": 1086976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.219461884230684e-05,
+      "loss": 3.9542,
+      "step": 1087488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.218623289479632e-05,
+      "loss": 3.9607,
+      "step": 1088000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.21778469472858e-05,
+      "loss": 3.9575,
+      "step": 1088512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.216946099977528e-05,
+      "loss": 3.9546,
+      "step": 1089024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.216107505226476e-05,
+      "loss": 3.961,
+      "step": 1089536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.215268910475425e-05,
+      "loss": 3.965,
+      "step": 1090048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.214431953604746e-05,
+      "loss": 3.9706,
+      "step": 1090560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.213593358853694e-05,
+      "loss": 3.9636,
+      "step": 1091072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.212754764102642e-05,
+      "loss": 3.9603,
+      "step": 1091584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.21191616935159e-05,
+      "loss": 3.9575,
+      "step": 1092096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.211077574600537e-05,
+      "loss": 3.9704,
+      "step": 1092608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.210238979849485e-05,
+      "loss": 3.9416,
+      "step": 1093120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.209400385098433e-05,
+      "loss": 3.9622,
+      "step": 1093632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.208561790347381e-05,
+      "loss": 3.9654,
+      "step": 1094144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.207724833476702e-05,
+      "loss": 3.9707,
+      "step": 1094656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.20688623872565e-05,
+      "loss": 3.9581,
+      "step": 1095168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.206047643974599e-05,
+      "loss": 3.9547,
+      "step": 1095680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.205209049223547e-05,
+      "loss": 3.9598,
+      "step": 1096192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.204370454472495e-05,
+      "loss": 3.9513,
+      "step": 1096704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.203531859721443e-05,
+      "loss": 3.9727,
+      "step": 1097216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2026949028507636e-05,
+      "loss": 3.9571,
+      "step": 1097728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2018563080997116e-05,
+      "loss": 3.9783,
+      "step": 1098240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2010177133486596e-05,
+      "loss": 3.9617,
+      "step": 1098752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.2001791185976076e-05,
+      "loss": 3.9437,
+      "step": 1099264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1993405238465556e-05,
+      "loss": 3.9754,
+      "step": 1099776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1985019290955036e-05,
+      "loss": 3.9513,
+      "step": 1100288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1976633343444516e-05,
+      "loss": 3.9504,
+      "step": 1100800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1968247395933996e-05,
+      "loss": 3.9618,
+      "step": 1101312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1959861448423476e-05,
+      "loss": 3.9651,
+      "step": 1101824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1951491879716685e-05,
+      "loss": 3.9485,
+      "step": 1102336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.194310593220617e-05,
+      "loss": 3.9542,
+      "step": 1102848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.193471998469565e-05,
+      "loss": 3.9496,
+      "step": 1103360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.192633403718513e-05,
+      "loss": 3.9487,
+      "step": 1103872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.191796446847834e-05,
+      "loss": 3.9694,
+      "step": 1104384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.190957852096782e-05,
+      "loss": 3.9624,
+      "step": 1104896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.19011925734573e-05,
+      "loss": 3.9565,
+      "step": 1105408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.189280662594678e-05,
+      "loss": 3.9574,
+      "step": 1105920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.188442067843626e-05,
+      "loss": 3.9685,
+      "step": 1106432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.187605110972947e-05,
+      "loss": 3.9602,
+      "step": 1106944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.186768154102268e-05,
+      "loss": 3.9628,
+      "step": 1107456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.185929559351216e-05,
+      "loss": 3.9563,
+      "step": 1107968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.185090964600164e-05,
+      "loss": 3.9588,
+      "step": 1108480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1842523698491126e-05,
+      "loss": 3.9641,
+      "step": 1108992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1834137750980606e-05,
+      "loss": 3.9588,
+      "step": 1109504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1825751803470086e-05,
+      "loss": 3.9573,
+      "step": 1110016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.181736585595956e-05,
+      "loss": 3.9658,
+      "step": 1110528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.180897990844904e-05,
+      "loss": 3.9631,
+      "step": 1111040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1800610339742255e-05,
+      "loss": 3.9455,
+      "step": 1111552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1792224392231735e-05,
+      "loss": 3.9555,
+      "step": 1112064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.178383844472121e-05,
+      "loss": 3.9679,
+      "step": 1112576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1775468876014424e-05,
+      "loss": 3.9527,
+      "step": 1113088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1767082928503904e-05,
+      "loss": 3.9585,
+      "step": 1113600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.175869698099338e-05,
+      "loss": 3.952,
+      "step": 1114112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1750311033482863e-05,
+      "loss": 3.9558,
+      "step": 1114624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.174194146477608e-05,
+      "loss": 3.9564,
+      "step": 1115136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.173355551726556e-05,
+      "loss": 3.9549,
+      "step": 1115648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.172516956975503e-05,
+      "loss": 3.959,
+      "step": 1116160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.171678362224451e-05,
+      "loss": 3.9589,
+      "step": 1116672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.170839767473399e-05,
+      "loss": 3.9715,
+      "step": 1117184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.170001172722347e-05,
+      "loss": 3.9351,
+      "step": 1117696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.169162577971295e-05,
+      "loss": 3.9576,
+      "step": 1118208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.168323983220243e-05,
+      "loss": 3.953,
+      "step": 1118720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.167485388469191e-05,
+      "loss": 3.9637,
+      "step": 1119232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.166650069478885e-05,
+      "loss": 3.9661,
+      "step": 1119744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.165811474727833e-05,
+      "loss": 3.9551,
+      "step": 1120256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.164972879976782e-05,
+      "loss": 3.951,
+      "step": 1120768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.164135923106103e-05,
+      "loss": 3.9564,
+      "step": 1121280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1632973283550506e-05,
+      "loss": 3.9555,
+      "step": 1121792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1624587336039986e-05,
+      "loss": 3.948,
+      "step": 1122304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1616201388529466e-05,
+      "loss": 3.9488,
+      "step": 1122816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1607815441018946e-05,
+      "loss": 3.9557,
+      "step": 1123328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1599429493508426e-05,
+      "loss": 3.9556,
+      "step": 1123840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1591043545997906e-05,
+      "loss": 3.9441,
+      "step": 1124352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1582657598487386e-05,
+      "loss": 3.9539,
+      "step": 1124864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1574271650976866e-05,
+      "loss": 3.9418,
+      "step": 1125376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1565885703466346e-05,
+      "loss": 3.9605,
+      "step": 1125888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1557499755955826e-05,
+      "loss": 3.9469,
+      "step": 1126400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1549113808445306e-05,
+      "loss": 3.952,
+      "step": 1126912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1540744239738515e-05,
+      "loss": 3.9571,
+      "step": 1127424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1532358292227995e-05,
+      "loss": 3.9598,
+      "step": 1127936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.152398872352121e-05,
+      "loss": 3.9531,
+      "step": 1128448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.151560277601069e-05,
+      "loss": 3.9504,
+      "step": 1128960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.15072332073039e-05,
+      "loss": 3.9606,
+      "step": 1129472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.149884725979338e-05,
+      "loss": 3.9491,
+      "step": 1129984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.149046131228286e-05,
+      "loss": 3.9459,
+      "step": 1130496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.148207536477234e-05,
+      "loss": 3.9583,
+      "step": 1131008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.147370579606555e-05,
+      "loss": 3.9572,
+      "step": 1131520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.146531984855503e-05,
+      "loss": 3.9572,
+      "step": 1132032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.145693390104451e-05,
+      "loss": 3.9483,
+      "step": 1132544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.144854795353399e-05,
+      "loss": 3.9519,
+      "step": 1133056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.144016200602347e-05,
+      "loss": 3.9494,
+      "step": 1133568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.143177605851295e-05,
+      "loss": 3.9463,
+      "step": 1134080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1423390111002435e-05,
+      "loss": 3.9421,
+      "step": 1134592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1415004163491915e-05,
+      "loss": 3.9643,
+      "step": 1135104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1406618215981395e-05,
+      "loss": 3.9632,
+      "step": 1135616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.139823226847087e-05,
+      "loss": 3.9507,
+      "step": 1136128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1389862699764084e-05,
+      "loss": 3.9505,
+      "step": 1136640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1381476752253564e-05,
+      "loss": 3.9592,
+      "step": 1137152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.137309080474304e-05,
+      "loss": 3.9534,
+      "step": 1137664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.136470485723252e-05,
+      "loss": 3.9511,
+      "step": 1138176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.135633528852573e-05,
+      "loss": 3.9536,
+      "step": 1138688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.134794934101521e-05,
+      "loss": 3.9644,
+      "step": 1139200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.1339563393504686e-05,
+      "loss": 3.9583,
+      "step": 1139712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.133117744599417e-05,
+      "loss": 3.9503,
+      "step": 1140224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.132279149848365e-05,
+      "loss": 3.9476,
+      "step": 1140736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.131440555097313e-05,
+      "loss": 3.9558,
+      "step": 1141248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.130601960346261e-05,
+      "loss": 3.9581,
+      "step": 1141760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.129763365595209e-05,
+      "loss": 3.9434,
+      "step": 1142272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.12892640872453e-05,
+      "loss": 3.9672,
+      "step": 1142784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.128087813973478e-05,
+      "loss": 3.9466,
+      "step": 1143296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.127249219222426e-05,
+      "loss": 3.9607,
+      "step": 1143808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.126410624471374e-05,
+      "loss": 3.9551,
+      "step": 1144320
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.069850444793701,
+      "eval_runtime": 297.2128,
+      "eval_samples_per_second": 1283.898,
+      "eval_steps_per_second": 40.123,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125573667600695e-05,
+      "loss": 3.9504,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124735072849643e-05,
+      "loss": 3.9495,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123896478098591e-05,
+      "loss": 3.955,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123057883347539e-05,
+      "loss": 3.9625,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122219288596487e-05,
+      "loss": 3.955,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121380693845436e-05,
+      "loss": 3.9558,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120542099094384e-05,
+      "loss": 3.9499,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119703504343332e-05,
+      "loss": 3.9502,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11886490959228e-05,
+      "loss": 3.9489,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118026314841228e-05,
+      "loss": 3.9598,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117187720090176e-05,
+      "loss": 3.9503,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116349125339123e-05,
+      "loss": 3.9547,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1155121684684446e-05,
+      "loss": 3.9639,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1146735737173926e-05,
+      "loss": 3.9438,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11383497896634e-05,
+      "loss": 3.9503,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112996384215288e-05,
+      "loss": 3.9333,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112157789464236e-05,
+      "loss": 3.9448,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111319194713184e-05,
+      "loss": 3.9528,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1104805999621326e-05,
+      "loss": 3.9449,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1096420052110806e-05,
+      "loss": 3.9529,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1088034104600286e-05,
+      "loss": 3.9684,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1079664535893495e-05,
+      "loss": 3.9495,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1071294967186704e-05,
+      "loss": 3.954,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1062909019676184e-05,
+      "loss": 3.9548,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1054523072165664e-05,
+      "loss": 3.9467,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1046137124655144e-05,
+      "loss": 3.9533,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1037751177144624e-05,
+      "loss": 3.9411,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1029365229634104e-05,
+      "loss": 3.9511,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1020979282123584e-05,
+      "loss": 3.9494,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1012593334613064e-05,
+      "loss": 3.9374,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100422376590627e-05,
+      "loss": 3.9504,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099583781839576e-05,
+      "loss": 3.9454,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098745187088524e-05,
+      "loss": 3.9564,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097906592337472e-05,
+      "loss": 3.9556,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09706799758642e-05,
+      "loss": 3.9423,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096229402835368e-05,
+      "loss": 3.9464,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095390808084316e-05,
+      "loss": 3.965,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094553851213637e-05,
+      "loss": 3.9442,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093715256462585e-05,
+      "loss": 3.9493,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092876661711533e-05,
+      "loss": 3.945,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092038066960481e-05,
+      "loss": 3.9437,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091199472209429e-05,
+      "loss": 3.9463,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090360877458376e-05,
+      "loss": 3.9619,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089523920587698e-05,
+      "loss": 3.954,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088685325836646e-05,
+      "loss": 3.9512,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0878467310855944e-05,
+      "loss": 3.9533,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087008136334542e-05,
+      "loss": 3.9425,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08616954158349e-05,
+      "loss": 3.9592,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085330946832438e-05,
+      "loss": 3.9284,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084492352081386e-05,
+      "loss": 3.9506,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0836553952107066e-05,
+      "loss": 3.9539,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0828168004596546e-05,
+      "loss": 3.9586,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0819782057086026e-05,
+      "loss": 3.9484,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811396109575506e-05,
+      "loss": 3.9387,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0803010162064986e-05,
+      "loss": 3.9498,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794624214554466e-05,
+      "loss": 3.936,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078625464584768e-05,
+      "loss": 3.9648,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077786869833716e-05,
+      "loss": 3.9417,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076948275082664e-05,
+      "loss": 3.9657,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076109680331612e-05,
+      "loss": 3.9467,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07527108558056e-05,
+      "loss": 3.9401,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074432490829508e-05,
+      "loss": 3.9561,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073593896078456e-05,
+      "loss": 3.939,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072755301327404e-05,
+      "loss": 3.9425,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071918344456725e-05,
+      "loss": 3.9446,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071079749705673e-05,
+      "loss": 3.9541,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070241154954621e-05,
+      "loss": 3.9376,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069402560203569e-05,
+      "loss": 3.9442,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068563965452517e-05,
+      "loss": 3.9353,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0677286464622115e-05,
+      "loss": 3.9372,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0668900517111595e-05,
+      "loss": 3.9598,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0660514569601075e-05,
+      "loss": 3.9544,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0652128622090555e-05,
+      "loss": 3.9434,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0643742674580035e-05,
+      "loss": 3.9476,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0635373105873244e-05,
+      "loss": 3.9544,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0626987158362724e-05,
+      "loss": 3.9502,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0618601210852204e-05,
+      "loss": 3.9488,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0610215263341684e-05,
+      "loss": 3.9479,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0601829315831164e-05,
+      "loss": 3.9467,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0593443368320644e-05,
+      "loss": 3.952,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0585057420810124e-05,
+      "loss": 3.9456,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0576671473299604e-05,
+      "loss": 3.9448,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056830190459281e-05,
+      "loss": 3.9563,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05599159570823e-05,
+      "loss": 3.9491,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055153000957178e-05,
+      "loss": 3.9335,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054314406206125e-05,
+      "loss": 3.9479,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053477449335447e-05,
+      "loss": 3.951,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052640492464768e-05,
+      "loss": 3.9446,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051801897713716e-05,
+      "loss": 3.9473,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509633029626638e-05,
+      "loss": 3.9397,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501247082116118e-05,
+      "loss": 3.9462,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.049287751340933e-05,
+      "loss": 3.943,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048450794470254e-05,
+      "loss": 3.9446,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.047612199719202e-05,
+      "loss": 3.9446,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.04677360496815e-05,
+      "loss": 3.9456,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045935010217098e-05,
+      "loss": 3.9617,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045096415466046e-05,
+      "loss": 3.9302,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442578207149943e-05,
+      "loss": 3.9379,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434192259639423e-05,
+      "loss": 3.9467,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0425806312128896e-05,
+      "loss": 3.9476,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0417420364618376e-05,
+      "loss": 3.9526,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0409034417107856e-05,
+      "loss": 3.9461,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040066484840107e-05,
+      "loss": 3.9376,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392278900890548e-05,
+      "loss": 3.9461,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0383892953380028e-05,
+      "loss": 3.9448,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375507005869508e-05,
+      "loss": 3.9356,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0367137437162717e-05,
+      "loss": 3.9357,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358751489652197e-05,
+      "loss": 3.9464,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0350381920945413e-05,
+      "loss": 3.9481,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0342012352238625e-05,
+      "loss": 3.9326,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0333626404728105e-05,
+      "loss": 3.9396,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0325240457217585e-05,
+      "loss": 3.931,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0316854509707065e-05,
+      "loss": 3.9485,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308468562196545e-05,
+      "loss": 3.9378,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0300082614686022e-05,
+      "loss": 3.942,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291696667175502e-05,
+      "loss": 3.9461,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028331071966498e-05,
+      "loss": 3.9475,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.027492477215446e-05,
+      "loss": 3.9422,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026653882464394e-05,
+      "loss": 3.9417,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025815287713342e-05,
+      "loss": 3.9481,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.02497669296229e-05,
+      "loss": 3.9405,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.024138098211238e-05,
+      "loss": 3.9343,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0233011413405594e-05,
+      "loss": 3.9462,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224625465895074e-05,
+      "loss": 3.9405,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0216255897188283e-05,
+      "loss": 3.9505,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0207869949677763e-05,
+      "loss": 3.9361,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0199484002167243e-05,
+      "loss": 3.9423,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0191098054656726e-05,
+      "loss": 3.9367,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182712107146206e-05,
+      "loss": 3.9398,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174326159635686e-05,
+      "loss": 3.9295,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0165940212125166e-05,
+      "loss": 3.9509,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0157554264614646e-05,
+      "loss": 3.9559,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0149168317104126e-05,
+      "loss": 3.9419,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0140782369593606e-05,
+      "loss": 3.9361,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0132396422083083e-05,
+      "loss": 3.9477,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0124010474572563e-05,
+      "loss": 3.9442,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0115624527062042e-05,
+      "loss": 3.9378,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010725495835526e-05,
+      "loss": 3.9429,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009886901084473e-05,
+      "loss": 3.9519,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090499442137947e-05,
+      "loss": 3.9458,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0082113494627427e-05,
+      "loss": 3.941,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073727547116907e-05,
+      "loss": 3.9365,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065341599606384e-05,
+      "loss": 3.9462,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0056955652095864e-05,
+      "loss": 3.9442,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004858608338908e-05,
+      "loss": 3.9322,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0040200135878553e-05,
+      "loss": 3.9566,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031814188368036e-05,
+      "loss": 3.9376,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023428240857516e-05,
+      "loss": 3.947,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015042293346996e-05,
+      "loss": 3.9493,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006672724640205e-05,
+      "loss": 3.9406,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.064294815063477,
+      "eval_runtime": 304.6171,
+      "eval_samples_per_second": 1252.691,
+      "eval_steps_per_second": 39.148,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9998286777129685e-05,
+      "loss": 3.9385,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9989900829619165e-05,
+      "loss": 3.9407,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9981514882108645e-05,
+      "loss": 3.953,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.997312893459813e-05,
+      "loss": 3.9431,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.996474298708761e-05,
+      "loss": 3.9463,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.995635703957709e-05,
+      "loss": 3.9362,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9947971092066568e-05,
+      "loss": 3.9425,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939585144556048e-05,
+      "loss": 3.9372,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9931215575849257e-05,
+      "loss": 3.9507,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9922829628338737e-05,
+      "loss": 3.9357,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.991444368082822e-05,
+      "loss": 3.9495,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.990609049092516e-05,
+      "loss": 3.9544,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989770454341464e-05,
+      "loss": 3.9338,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.988931859590412e-05,
+      "loss": 3.9365,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.98809326483936e-05,
+      "loss": 3.9275,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872546700883082e-05,
+      "loss": 3.9317,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9864160753372562e-05,
+      "loss": 3.9418,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9855774805862042e-05,
+      "loss": 3.934,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9847388858351522e-05,
+      "loss": 3.9392,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9839002910841002e-05,
+      "loss": 3.9615,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830616963330482e-05,
+      "loss": 3.9382,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9822231015819962e-05,
+      "loss": 3.9447,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9813861447113174e-05,
+      "loss": 3.9409,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805475499602654e-05,
+      "loss": 3.9367,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9797089552092134e-05,
+      "loss": 3.9439,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9788703604581614e-05,
+      "loss": 3.9298,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9780334035874823e-05,
+      "loss": 3.936,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9771948088364303e-05,
+      "loss": 3.9413,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763562140853783e-05,
+      "loss": 3.9249,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9755176193343267e-05,
+      "loss": 3.9406,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.974679024583274e-05,
+      "loss": 3.9333,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.973840429832222e-05,
+      "loss": 3.944,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.97300183508117e-05,
+      "loss": 3.9404,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9721648782104916e-05,
+      "loss": 3.934,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9713262834594392e-05,
+      "loss": 3.938,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9704876887083872e-05,
+      "loss": 3.9543,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9696490939573352e-05,
+      "loss": 3.9305,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9688104992062832e-05,
+      "loss": 3.9395,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967973542335604e-05,
+      "loss": 3.9366,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967134947584552e-05,
+      "loss": 3.9264,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9662963528335004e-05,
+      "loss": 3.9356,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9654577580824484e-05,
+      "loss": 3.9497,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9646191633313964e-05,
+      "loss": 3.947,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9637805685803444e-05,
+      "loss": 3.9378,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9629436117096653e-05,
+      "loss": 3.9439,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9621050169586133e-05,
+      "loss": 3.934,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9612664222075613e-05,
+      "loss": 3.9471,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9604278274565093e-05,
+      "loss": 3.9219,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9595892327054577e-05,
+      "loss": 3.9374,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9587522758347786e-05,
+      "loss": 3.9465,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9579136810837266e-05,
+      "loss": 3.9439,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9570750863326746e-05,
+      "loss": 3.9386,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9562364915816225e-05,
+      "loss": 3.9279,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9553978968305705e-05,
+      "loss": 3.9406,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9545609399598918e-05,
+      "loss": 3.9246,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9537223452088398e-05,
+      "loss": 3.9522,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9528837504577878e-05,
+      "loss": 3.9308,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9520451557067358e-05,
+      "loss": 3.9517,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9512065609556838e-05,
+      "loss": 3.938,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9503696040850047e-05,
+      "loss": 3.932,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.949531009333953e-05,
+      "loss": 3.94,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.948692414582901e-05,
+      "loss": 3.9332,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.947853819831849e-05,
+      "loss": 3.9333,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.947015225080797e-05,
+      "loss": 3.9327,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.946176630329745e-05,
+      "loss": 3.9461,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.945338035578693e-05,
+      "loss": 3.9295,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.944501078708014e-05,
+      "loss": 3.9317,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9436624839569622e-05,
+      "loss": 3.9234,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9428238892059102e-05,
+      "loss": 3.9269,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9419852944548576e-05,
+      "loss": 3.9482,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9411466997038055e-05,
+      "loss": 3.9436,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9403081049527535e-05,
+      "loss": 3.934,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394695102017015e-05,
+      "loss": 3.9361,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.93863091545065e-05,
+      "loss": 3.9425,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9377939585799708e-05,
+      "loss": 3.9439,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369570017092924e-05,
+      "loss": 3.933,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9361184069582404e-05,
+      "loss": 3.9409,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352798122071877e-05,
+      "loss": 3.9361,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.934441217456136e-05,
+      "loss": 3.9422,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.933602622705084e-05,
+      "loss": 3.936,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.932764027954032e-05,
+      "loss": 3.9365,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.93192543320298e-05,
+      "loss": 3.9461,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.931088476332301e-05,
+      "loss": 3.9383,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.930249881581249e-05,
+      "loss": 3.923,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.929411286830197e-05,
+      "loss": 3.9367,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9285726920791452e-05,
+      "loss": 3.9403,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9277340973280932e-05,
+      "loss": 3.9375,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926897140457414e-05,
+      "loss": 3.937,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926058545706362e-05,
+      "loss": 3.9295,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.92521995095531e-05,
+      "loss": 3.9368,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.924381356204258e-05,
+      "loss": 3.9296,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235443993335794e-05,
+      "loss": 3.936,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9227058045825274e-05,
+      "loss": 3.9331,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9218672098314754e-05,
+      "loss": 3.9338,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9210286150804234e-05,
+      "loss": 3.9553,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9201900203293714e-05,
+      "loss": 3.9232,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9193514255783194e-05,
+      "loss": 3.9228,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9185128308272674e-05,
+      "loss": 3.9369,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9176742360762153e-05,
+      "loss": 3.9394,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9168372792055366e-05,
+      "loss": 3.9459,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9160003223348575e-05,
+      "loss": 3.9335,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9151617275838055e-05,
+      "loss": 3.9273,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9143231328327535e-05,
+      "loss": 3.9323,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9134845380817015e-05,
+      "loss": 3.9404,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9126459433306498e-05,
+      "loss": 3.9245,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9118089864599707e-05,
+      "loss": 3.9247,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9109703917089187e-05,
+      "loss": 3.9379,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9101317969578667e-05,
+      "loss": 3.9334,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9092932022068147e-05,
+      "loss": 3.9237,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9084546074557627e-05,
+      "loss": 3.9273,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9076160127047107e-05,
+      "loss": 3.9252,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9067774179536587e-05,
+      "loss": 3.939,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9059388232026064e-05,
+      "loss": 3.927,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9051002284515544e-05,
+      "loss": 3.9328,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.904263271580876e-05,
+      "loss": 3.9349,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9034246768298236e-05,
+      "loss": 3.9402,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.902587719959145e-05,
+      "loss": 3.9291,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9017491252080932e-05,
+      "loss": 3.9335,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9009105304570412e-05,
+      "loss": 3.9374,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9000719357059885e-05,
+      "loss": 3.9337,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8992333409549365e-05,
+      "loss": 3.9165,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8983947462038845e-05,
+      "loss": 3.9375,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.897557789333206e-05,
+      "loss": 3.9327,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8967191945821537e-05,
+      "loss": 3.939,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8958805998311017e-05,
+      "loss": 3.9231,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950420050800497e-05,
+      "loss": 3.9363,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8942034103289977e-05,
+      "loss": 3.9261,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933648155779457e-05,
+      "loss": 3.9299,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.892527858707267e-05,
+      "loss": 3.9139,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.891689263956215e-05,
+      "loss": 3.9429,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.890850669205163e-05,
+      "loss": 3.9448,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.890013712334484e-05,
+      "loss": 3.9324,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.889175117583432e-05,
+      "loss": 3.9221,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88833652283238e-05,
+      "loss": 3.9429,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.887497928081328e-05,
+      "loss": 3.9335,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8866593333302762e-05,
+      "loss": 3.9336,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8858207385792242e-05,
+      "loss": 3.9331,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8849821438281722e-05,
+      "loss": 3.9416,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8841435490771202e-05,
+      "loss": 3.9331,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8833049543260682e-05,
+      "loss": 3.9332,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8824663595750162e-05,
+      "loss": 3.9249,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.881627764823964e-05,
+      "loss": 3.9383,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8807908079532854e-05,
+      "loss": 3.9339,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8799522132022334e-05,
+      "loss": 3.9249,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8791152563315543e-05,
+      "loss": 3.9473,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8782766615805023e-05,
+      "loss": 3.9257,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8774380668294503e-05,
+      "loss": 3.9358,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8765994720783983e-05,
+      "loss": 3.9398,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8757608773273463e-05,
+      "loss": 3.9305,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.059375762939453,
+      "eval_runtime": 305.6668,
+      "eval_samples_per_second": 1248.389,
+      "eval_steps_per_second": 39.013,
+      "step": 1297440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8749222825762946e-05,
+      "loss": 3.9272,
+      "step": 1297920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8740836878252426e-05,
+      "loss": 3.9336,
+      "step": 1298432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.87324509307419e-05,
+      "loss": 3.9426,
+      "step": 1298944
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8724081362035115e-05,
+      "loss": 3.9339,
+      "step": 1299456
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8715695414524595e-05,
+      "loss": 3.942,
+      "step": 1299968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8707309467014072e-05,
+      "loss": 3.9247,
+      "step": 1300480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8698923519503552e-05,
+      "loss": 3.9315,
+      "step": 1300992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8690553950796768e-05,
+      "loss": 3.9244,
+      "step": 1301504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8682184382089977e-05,
+      "loss": 3.9426,
+      "step": 1302016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8673798434579457e-05,
+      "loss": 3.9283,
+      "step": 1302528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.866542886587267e-05,
+      "loss": 3.936,
+      "step": 1303040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.865704291836215e-05,
+      "loss": 3.9413,
+      "step": 1303552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.864867334965536e-05,
+      "loss": 3.9254,
+      "step": 1304064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8640287402144838e-05,
+      "loss": 3.9266,
+      "step": 1304576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8631901454634318e-05,
+      "loss": 3.918,
+      "step": 1305088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8623515507123798e-05,
+      "loss": 3.9268,
+      "step": 1305600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8615129559613278e-05,
+      "loss": 3.9272,
+      "step": 1306112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.860674361210276e-05,
+      "loss": 3.9268,
+      "step": 1306624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.859835766459224e-05,
+      "loss": 3.9252,
+      "step": 1307136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.858997171708172e-05,
+      "loss": 3.9516,
+      "step": 1307648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8581585769571195e-05,
+      "loss": 3.9304,
+      "step": 1308160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8573199822060675e-05,
+      "loss": 3.9321,
+      "step": 1308672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8564813874550154e-05,
+      "loss": 3.936,
+      "step": 1309184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8556427927039638e-05,
+      "loss": 3.931,
+      "step": 1309696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8548041979529118e-05,
+      "loss": 3.9281,
+      "step": 1310208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8539656032018598e-05,
+      "loss": 3.9225,
+      "step": 1310720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8531270084508078e-05,
+      "loss": 3.9295,
+      "step": 1311232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8522884136997558e-05,
+      "loss": 3.9302,
+      "step": 1311744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8514514568290767e-05,
+      "loss": 3.9159,
+      "step": 1312256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8506128620780247e-05,
+      "loss": 3.9316,
+      "step": 1312768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.849775905207346e-05,
+      "loss": 3.9217,
+      "step": 1313280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.848937310456294e-05,
+      "loss": 3.9357,
+      "step": 1313792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.848098715705242e-05,
+      "loss": 3.934,
+      "step": 1314304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.84726012095419e-05,
+      "loss": 3.9207,
+      "step": 1314816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.846421526203138e-05,
+      "loss": 3.9327,
+      "step": 1315328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.845584569332459e-05,
+      "loss": 3.9413,
+      "step": 1315840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.84474761246178e-05,
+      "loss": 3.918,
+      "step": 1316352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.843909017710728e-05,
+      "loss": 3.9334,
+      "step": 1316864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.843070422959676e-05,
+      "loss": 3.9245,
+      "step": 1317376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.842233466088997e-05,
+      "loss": 3.9197,
+      "step": 1317888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8413948713379453e-05,
+      "loss": 3.9295,
+      "step": 1318400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8405562765868933e-05,
+      "loss": 3.9381,
+      "step": 1318912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8397176818358413e-05,
+      "loss": 3.9327,
+      "step": 1319424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8388790870847893e-05,
+      "loss": 3.9271,
+      "step": 1319936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8380404923337373e-05,
+      "loss": 3.9391,
+      "step": 1320448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8372018975826853e-05,
+      "loss": 3.9221,
+      "step": 1320960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8363633028316333e-05,
+      "loss": 3.9345,
+      "step": 1321472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8355247080805813e-05,
+      "loss": 3.9129,
+      "step": 1321984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8346861133295293e-05,
+      "loss": 3.9226,
+      "step": 1322496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8338475185784773e-05,
+      "loss": 3.9362,
+      "step": 1323008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8330089238274256e-05,
+      "loss": 3.936,
+      "step": 1323520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.832170329076373e-05,
+      "loss": 3.9246,
+      "step": 1324032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8313333722056945e-05,
+      "loss": 3.9182,
+      "step": 1324544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8304947774546425e-05,
+      "loss": 3.933,
+      "step": 1325056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8296561827035905e-05,
+      "loss": 3.9141,
+      "step": 1325568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8288192258329117e-05,
+      "loss": 3.94,
+      "step": 1326080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8279806310818597e-05,
+      "loss": 3.9236,
+      "step": 1326592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8271420363308077e-05,
+      "loss": 3.9423,
+      "step": 1327104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8263050794601286e-05,
+      "loss": 3.9283,
+      "step": 1327616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8254664847090766e-05,
+      "loss": 3.9232,
+      "step": 1328128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8246278899580246e-05,
+      "loss": 3.9253,
+      "step": 1328640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8237892952069726e-05,
+      "loss": 3.9319,
+      "step": 1329152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8229507004559203e-05,
+      "loss": 3.918,
+      "step": 1329664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8221121057048683e-05,
+      "loss": 3.9229,
+      "step": 1330176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8212735109538163e-05,
+      "loss": 3.9365,
+      "step": 1330688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8204349162027643e-05,
+      "loss": 3.9239,
+      "step": 1331200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8195963214517123e-05,
+      "loss": 3.9173,
+      "step": 1331712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8187577267006603e-05,
+      "loss": 3.9153,
+      "step": 1332224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8179191319496086e-05,
+      "loss": 3.9173,
+      "step": 1332736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8170805371985566e-05,
+      "loss": 3.9375,
+      "step": 1333248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8162435803278775e-05,
+      "loss": 3.9351,
+      "step": 1333760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8154066234571984e-05,
+      "loss": 3.9233,
+      "step": 1334272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8145680287061464e-05,
+      "loss": 3.9275,
+      "step": 1334784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8137294339550947e-05,
+      "loss": 3.9311,
+      "step": 1335296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8128908392040427e-05,
+      "loss": 3.9307,
+      "step": 1335808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8120538823333636e-05,
+      "loss": 3.9226,
+      "step": 1336320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8112152875823116e-05,
+      "loss": 3.9316,
+      "step": 1336832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8103766928312596e-05,
+      "loss": 3.927,
+      "step": 1337344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8095380980802076e-05,
+      "loss": 3.9374,
+      "step": 1337856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8086995033291556e-05,
+      "loss": 3.9246,
+      "step": 1338368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.807860908578104e-05,
+      "loss": 3.9278,
+      "step": 1338880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.807022313827052e-05,
+      "loss": 3.9333,
+      "step": 1339392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.806183719076e-05,
+      "loss": 3.9303,
+      "step": 1339904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.805345124324948e-05,
+      "loss": 3.9148,
+      "step": 1340416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.804506529573896e-05,
+      "loss": 3.9269,
+      "step": 1340928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.803667934822844e-05,
+      "loss": 3.9264,
+      "step": 1341440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8028293400717916e-05,
+      "loss": 3.9302,
+      "step": 1341952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8019923832011132e-05,
+      "loss": 3.9286,
+      "step": 1342464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.801155426330434e-05,
+      "loss": 3.9183,
+      "step": 1342976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.800316831579382e-05,
+      "loss": 3.9259,
+      "step": 1343488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.79947823682833e-05,
+      "loss": 3.9218,
+      "step": 1344000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.798639642077278e-05,
+      "loss": 3.9312,
+      "step": 1344512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.797801047326226e-05,
+      "loss": 3.9163,
+      "step": 1345024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7969640904555473e-05,
+      "loss": 3.9251,
+      "step": 1345536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7961254957044953e-05,
+      "loss": 3.9388,
+      "step": 1346048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7952869009534433e-05,
+      "loss": 3.9265,
+      "step": 1346560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7944483062023913e-05,
+      "loss": 3.907,
+      "step": 1347072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7936097114513386e-05,
+      "loss": 3.9267,
+      "step": 1347584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.792771116700287e-05,
+      "loss": 3.9309,
+      "step": 1348096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.791932521949235e-05,
+      "loss": 3.933,
+      "step": 1348608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.791093927198183e-05,
+      "loss": 3.9231,
+      "step": 1349120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.790255332447131e-05,
+      "loss": 3.9199,
+      "step": 1349632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.789418375576452e-05,
+      "loss": 3.9181,
+      "step": 1350144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7885797808254e-05,
+      "loss": 3.9318,
+      "step": 1350656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.787741186074348e-05,
+      "loss": 3.9108,
+      "step": 1351168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.786904229203669e-05,
+      "loss": 3.9186,
+      "step": 1351680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.786065634452617e-05,
+      "loss": 3.9263,
+      "step": 1352192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.785227039701565e-05,
+      "loss": 3.9295,
+      "step": 1352704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.784388444950513e-05,
+      "loss": 3.9087,
+      "step": 1353216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.783549850199461e-05,
+      "loss": 3.9235,
+      "step": 1353728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.782711255448409e-05,
+      "loss": 3.9128,
+      "step": 1354240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7818742985777303e-05,
+      "loss": 3.929,
+      "step": 1354752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7810357038266783e-05,
+      "loss": 3.9181,
+      "step": 1355264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7801971090756263e-05,
+      "loss": 3.9137,
+      "step": 1355776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7793585143245743e-05,
+      "loss": 3.9342,
+      "step": 1356288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7785215574538952e-05,
+      "loss": 3.9285,
+      "step": 1356800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7776846005832165e-05,
+      "loss": 3.9212,
+      "step": 1357312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7768460058321645e-05,
+      "loss": 3.9255,
+      "step": 1357824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7760074110811125e-05,
+      "loss": 3.9229,
+      "step": 1358336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7751688163300605e-05,
+      "loss": 3.9272,
+      "step": 1358848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7743302215790084e-05,
+      "loss": 3.9057,
+      "step": 1359360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7734916268279564e-05,
+      "loss": 3.9299,
+      "step": 1359872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7726530320769044e-05,
+      "loss": 3.9202,
+      "step": 1360384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7718144373258524e-05,
+      "loss": 3.9284,
+      "step": 1360896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7709758425748004e-05,
+      "loss": 3.9172,
+      "step": 1361408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7701372478237488e-05,
+      "loss": 3.9233,
+      "step": 1361920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7693002909530697e-05,
+      "loss": 3.9162,
+      "step": 1362432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7684616962020177e-05,
+      "loss": 3.9217,
+      "step": 1362944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7676231014509657e-05,
+      "loss": 3.9047,
+      "step": 1363456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7667845066999137e-05,
+      "loss": 3.9363,
+      "step": 1363968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.765947549829235e-05,
+      "loss": 3.9303,
+      "step": 1364480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.765108955078183e-05,
+      "loss": 3.9244,
+      "step": 1364992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.764270360327131e-05,
+      "loss": 3.9167,
+      "step": 1365504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.763431765576079e-05,
+      "loss": 3.9295,
+      "step": 1366016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.762593170825027e-05,
+      "loss": 3.9291,
+      "step": 1366528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7617562139543478e-05,
+      "loss": 3.92,
+      "step": 1367040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7609176192032958e-05,
+      "loss": 3.9192,
+      "step": 1367552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.760079024452244e-05,
+      "loss": 3.9306,
+      "step": 1368064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.759240429701192e-05,
+      "loss": 3.9215,
+      "step": 1368576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.75840183495014e-05,
+      "loss": 3.93,
+      "step": 1369088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7575632401990874e-05,
+      "loss": 3.9141,
+      "step": 1369600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7567246454480354e-05,
+      "loss": 3.9279,
+      "step": 1370112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7558860506969834e-05,
+      "loss": 3.9245,
+      "step": 1370624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.755049093826305e-05,
+      "loss": 3.916,
+      "step": 1371136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7542121369556263e-05,
+      "loss": 3.9375,
+      "step": 1371648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7533735422045743e-05,
+      "loss": 3.9174,
+      "step": 1372160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7525349474535223e-05,
+      "loss": 3.9255,
+      "step": 1372672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.751697990582843e-05,
+      "loss": 3.9273,
+      "step": 1373184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.750859395831791e-05,
+      "loss": 3.9228,
+      "step": 1373696
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.055521011352539,
+      "eval_runtime": 293.5471,
+      "eval_samples_per_second": 1299.931,
+      "eval_steps_per_second": 40.624,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500208010807395e-05,
+      "loss": 3.9175,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491822063296875e-05,
+      "loss": 3.9243,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483436115786348e-05,
+      "loss": 3.9295,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7475050168275828e-05,
+      "loss": 3.9259,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466664220765308e-05,
+      "loss": 3.9331,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458278273254788e-05,
+      "loss": 3.9143,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744989232574427e-05,
+      "loss": 3.9202,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744150637823375e-05,
+      "loss": 3.9178,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743312043072323e-05,
+      "loss": 3.9309,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742473448321271e-05,
+      "loss": 3.9205,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741634853570219e-05,
+      "loss": 3.9282,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740796258819167e-05,
+      "loss": 3.9309,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739959301948488e-05,
+      "loss": 3.917,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391207071974364e-05,
+      "loss": 3.9195,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382821124463843e-05,
+      "loss": 3.9075,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374435176953323e-05,
+      "loss": 3.9159,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7366049229442803e-05,
+      "loss": 3.9197,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357663281932283e-05,
+      "loss": 3.9163,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349277334421763e-05,
+      "loss": 3.9176,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734089138691124e-05,
+      "loss": 3.9421,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.733250543940072e-05,
+      "loss": 3.9218,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324135870693936e-05,
+      "loss": 3.9217,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731574992318341e-05,
+      "loss": 3.9272,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.730736397567289e-05,
+      "loss": 3.9236,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729897802816237e-05,
+      "loss": 3.9177,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290608459455585e-05,
+      "loss": 3.9118,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282238890748797e-05,
+      "loss": 3.9232,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273852943238277e-05,
+      "loss": 3.9176,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265466995727757e-05,
+      "loss": 3.9103,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7257081048217237e-05,
+      "loss": 3.923,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724869510070671e-05,
+      "loss": 3.9146,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724030915319619e-05,
+      "loss": 3.9246,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7231923205685673e-05,
+      "loss": 3.9215,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223537258175153e-05,
+      "loss": 3.9115,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7215151310664633e-05,
+      "loss": 3.9228,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7206765363154113e-05,
+      "loss": 3.9355,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7198379415643593e-05,
+      "loss": 3.9065,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7189993468133073e-05,
+      "loss": 3.9265,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7181623899426282e-05,
+      "loss": 3.9127,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173254330719495e-05,
+      "loss": 3.914,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164868383208975e-05,
+      "loss": 3.9161,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156482435698455e-05,
+      "loss": 3.9326,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7148112866991664e-05,
+      "loss": 3.9224,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139726919481144e-05,
+      "loss": 3.9206,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131340971970627e-05,
+      "loss": 3.9274,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122955024460107e-05,
+      "loss": 3.9137,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114569076949587e-05,
+      "loss": 3.9288,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106183129439067e-05,
+      "loss": 3.9074,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097797181928547e-05,
+      "loss": 3.913,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089411234418027e-05,
+      "loss": 3.9264,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7081041665711236e-05,
+      "loss": 3.9255,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.707265571820072e-05,
+      "loss": 3.9174,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.70642697706902e-05,
+      "loss": 3.9119,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.705588382317968e-05,
+      "loss": 3.9236,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704751425447289e-05,
+      "loss": 3.9061,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703912830696237e-05,
+      "loss": 3.929,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030742359451848e-05,
+      "loss": 3.919,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022356411941328e-05,
+      "loss": 3.9296,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701397046443081e-05,
+      "loss": 3.9249,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700558451692029e-05,
+      "loss": 3.9118,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.699719856940977e-05,
+      "loss": 3.9159,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6988812621899245e-05,
+      "loss": 3.9236,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6980426674388725e-05,
+      "loss": 3.9075,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697205710568194e-05,
+      "loss": 3.9143,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.696367115817142e-05,
+      "loss": 3.9237,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955285210660897e-05,
+      "loss": 3.918,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946899263150377e-05,
+      "loss": 3.9061,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938513315639857e-05,
+      "loss": 3.9074,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930127368129337e-05,
+      "loss": 3.9069,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921741420618817e-05,
+      "loss": 3.9279,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913355473108297e-05,
+      "loss": 3.9251,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.690498590440151e-05,
+      "loss": 3.9144,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689659995689099e-05,
+      "loss": 3.9225,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68882303881842e-05,
+      "loss": 3.919,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.687984444067368e-05,
+      "loss": 3.9263,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871458493163158e-05,
+      "loss": 3.9129,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.686307254565264e-05,
+      "loss": 3.9259,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.685468659814212e-05,
+      "loss": 3.9212,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68463006506316e-05,
+      "loss": 3.9235,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683791470312108e-05,
+      "loss": 3.9148,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682952875561056e-05,
+      "loss": 3.919,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682115918690377e-05,
+      "loss": 3.9218,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.681277323939325e-05,
+      "loss": 3.9221,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.680438729188273e-05,
+      "loss": 3.9109,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6796001344372214e-05,
+      "loss": 3.9178,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787631775665423e-05,
+      "loss": 3.918,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779245828154903e-05,
+      "loss": 3.9207,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770859880644383e-05,
+      "loss": 3.9239,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762473933133863e-05,
+      "loss": 3.9067,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6754087985623343e-05,
+      "loss": 3.918,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745718416916555e-05,
+      "loss": 3.9154,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737332469406035e-05,
+      "loss": 3.9237,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728946521895515e-05,
+      "loss": 3.9041,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720576953188724e-05,
+      "loss": 3.9185,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712191005678204e-05,
+      "loss": 3.9298,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703805058167684e-05,
+      "loss": 3.9177,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695419110657167e-05,
+      "loss": 3.8977,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6687033163146647e-05,
+      "loss": 3.9168,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678647215636127e-05,
+      "loss": 3.9225,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670261268125607e-05,
+      "loss": 3.9238,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661891699418816e-05,
+      "loss": 3.9172,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653505751908296e-05,
+      "loss": 3.9101,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645119804397776e-05,
+      "loss": 3.9107,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.663673385688726e-05,
+      "loss": 3.9213,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628347909376733e-05,
+      "loss": 3.9049,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661997834066995e-05,
+      "loss": 3.9104,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6611608771963158e-05,
+      "loss": 3.9176,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6603222824452638e-05,
+      "loss": 3.9195,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.659483687694212e-05,
+      "loss": 3.8991,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.65864509294316e-05,
+      "loss": 3.9168,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657806498192108e-05,
+      "loss": 3.9053,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6569679034410554e-05,
+      "loss": 3.9193,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6561293086900034e-05,
+      "loss": 3.9126,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552907139389514e-05,
+      "loss": 3.9068,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544521191878997e-05,
+      "loss": 3.923,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6536135244368477e-05,
+      "loss": 3.921,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527782054465422e-05,
+      "loss": 3.9097,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519396106954902e-05,
+      "loss": 3.9117,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6511010159444376e-05,
+      "loss": 3.9182,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.650262421193386e-05,
+      "loss": 3.9169,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.649423826442334e-05,
+      "loss": 3.9012,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6485868695716555e-05,
+      "loss": 3.9198,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477482748206028e-05,
+      "loss": 3.9137,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6469096800695508e-05,
+      "loss": 3.9176,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460710853184988e-05,
+      "loss": 3.913,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452324905674468e-05,
+      "loss": 3.918,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644393895816395e-05,
+      "loss": 3.9036,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643556938945716e-05,
+      "loss": 3.9123,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642718344194664e-05,
+      "loss": 3.8967,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641879749443612e-05,
+      "loss": 3.9269,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64104115469256e-05,
+      "loss": 3.9194,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.640202559941508e-05,
+      "loss": 3.9172,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639363965190456e-05,
+      "loss": 3.9051,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385253704394043e-05,
+      "loss": 3.9211,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376867756883523e-05,
+      "loss": 3.921,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368481809373003e-05,
+      "loss": 3.91,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360095861862483e-05,
+      "loss": 3.9121,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351709914351963e-05,
+      "loss": 3.9205,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343323966841443e-05,
+      "loss": 3.9143,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334954398134652e-05,
+      "loss": 3.9245,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326568450624136e-05,
+      "loss": 3.907,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318182503113616e-05,
+      "loss": 3.9175,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.630979655560309e-05,
+      "loss": 3.9118,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.630141060809257e-05,
+      "loss": 3.9064,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6293057418189514e-05,
+      "loss": 3.9308,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284671470678997e-05,
+      "loss": 3.9078,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276285523168477e-05,
+      "loss": 3.9149,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267899575657957e-05,
+      "loss": 3.9203,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259513628147437e-05,
+      "loss": 3.9128,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.051382541656494,
+      "eval_runtime": 299.6015,
+      "eval_samples_per_second": 1273.662,
+      "eval_steps_per_second": 39.803,
+      "step": 1450080
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 5.5850208657184115e+17,
+  "trial_name": null,
+  "trial_params": null
+}